How to Test a Car Battery in 2026: The Numbers Battery Makers Publish, and When It's the Alternator

This page answers one question: how to test your own car battery, what the numbers mean, and when the problem is the alternator instead. Every procedure and number comes from a document published under its own name by a battery, tester or alternator maker, a federal regulator, a retailer or, for the complaint figures, PulsCar's own study, read and saved on 30 September 2026, and where two makers print different numbers, both are shown. If your battery dies again and again, the drain hunt is in car battery keeps dying; if you already know you need one, prices are in car battery replacement cost.

The answer in 60 seconds: six tests, what each tells you, and the pass line

The order matters more than any single number. East Penn's sequence puts the visual check first, the voltage second and the load or conductance test third, and says the battery must be charged before the third step means anything. Denso, which makes alternators, says the same even for its own product: "A visual inspection and a performance test of the battery must always be performed before inspecting the charging system." Midtronics gives the reason: "The performance of the starting and charging systems depends on the battery’s condition."

What you need: a digital multimeter (or a tester's own voltage readout) for tests 2, 5 and 6, and either a load tester or a conductance tester for tests 3 and 4, or a free test at a parts store (below). What these tools cost and which kinds exist is covered in car diagnostic tools; this page quotes no prices.

Before you touch the battery: the safety rules the makers print

East Penn, Schumacher and Clarios print full safety warnings; Midtronics and Fluke print shorter cautions. Clarios, the company named on the VARTA and OPTIMA sites, says in its safety data sheet for lead acid batteries (revision dated 12/13/2023): "Highly flammable hydrogen gas is generated during charging and operation of batteries." If that gas is lit by a cigarette, a flame or a spark, the same sheet says it "may cause battery explosion with dispersion of casing fragments and corrosive liquid electrolyte."

Eyes first. East Penn, which makes Deka batteries, prints "ALWAYS WEAR EYE PROTECTION WHEN WORKING AROUND BATTERIES!" across its Basic Rules for Battery Care & Maintenance (Form 0004, dated 02/26) and asks for "safety glasses and a face shield". It adds: "Never lean directly over a battery while boosting, testing or charging it." Schumacher's BT-100 load tester manual adds: "Always have someone nearby for help."

No sparks near the battery. East Penn: "To avoid creating sparks, always turn charging and testing equipment off before attaching or removing clamps." And its rule for the battery's own cables: "ALWAYS DISCONNECT GROUNDED CABLE FIRST AND CONNECT IT LAST TO PREVENT DANGEROUS SPARKS."

No metal across the terminals, and no jewelry. Clarios says "do not allow metallic articles to simultaneously contact the negative and positive terminals of a battery." Schumacher explains why rings come off: "A lead-acid battery can produce a short circuit current high enough to weld a ring to metal, causing a severe burn."

If acid touches you. Clarios's sheet, for eyes: "Flush immediately with large amounts of water for at least 15 minutes while lifting lids; Seek immediate medical attention if eyes have been exposed directly to acid." Schumacher's manual says at least 10 minutes; Clarios's figure is the longer one. Schumacher says to keep water, soap and baking soda nearby, and East Penn neutralizes spills with a paste of baking soda and water.

When not to test at all. Schumacher: "Inspect the battery for a cracked or broken case or cover. If the battery is damaged, do not use the tester." OPTIMA, in its September 2026 guide to the signs a battery is dying: "A battery with a cracked case should be replaced immediately." Interstate's pro, on a battery that smells of rotten eggs: "this is a battery that is overcharging", followed by "So don’t go near that battery." And East Penn on a frozen battery: "To avoid a battery explosion, never attempt to charge a frozen battery."

What the federal workplace rules require. OSHA's rules are for workplaces, not driveways. Under 29 CFR 1926.441, "Face shields, aprons, and rubber gloves shall be provided for workers handling acids or batteries", and "Facilities for quick drenching of the eyes and body shall be provided within 25 feet (7.62 m) of battery handling areas." The industrial-truck rule, 29 CFR 1910.178(g), adds that metal tools stay off uncovered batteries.

And East Penn: "If you are uncertain about your ability to remove, install, or service a battery yourself do not attempt, and consult a professional!" A free parts-store test is a fair answer to this whole page.

Test 1: look before you measure

East Penn's in-vehicle procedure starts with the eyes, not the meter: "Prior to any testing, visually inspect the battery." Its list is five items: a cracked or broken case or cover, loose cable connections, a leaking case-to-cover seal, corrosion, and damaged or leaking terminals.

Bulging. Interstate's pro lists a battery "swollen like a football"; OPTIMA says "A battery that is swollen may continue to operate for a period of time, but replacement should be a priority."

Corrosion, and which terminal it is on. OPTIMA's guide reads the pattern: "If you only notice heavy corrosion on the positive terminal, it could indicate the battery is being overcharged. If you only notice heavy corrosion on the negative terminal, it could suggest the battery is being undercharged or there is a poor ground connection." Corrosion on its own does not mean the battery is bad, OPTIMA says, but it can cause "slow cranking, clicking or a no-start situation". East Penn's cleaning method: neutralize with a baking soda and water paste or a battery cleaner spray, brush off the residue and wash the area with clean water. Clean before you test.

Loose clamps and hold-downs. East Penn's marine brochure: tighten "snugly if appropriate", but "DO NOT OVER-TIGHTEN TO AVOID BATTERY DAMAGE!"

Age. Batteries carry a date code; East Penn's chart reads 9/26 as September 2026. Interstate's video says "a battery is at its best for three years". OPTIMA says a well-maintained battery not exposed to extreme temperatures can last "significantly longer" than the 3-5 year timespan it says is frequently offered. Age is a reason to test, not a verdict.

Where your battery is. If yours is in the trunk or cabin, VARTA's step-by-step testing guide says a tester must go on the battery's own terminals, "and not the starting aid contacts in the engine compartment", because the cable in between changes the reading.

Test 2: the resting voltage, step by step with a multimeter

A digital multimeter is the tool for tests 2, 5 and 6. Fluke's guide to measuring battery voltage (updated February 2026) gives the steps:

1. Everything off. Fluke: "Turn off all accessories before taking a resting voltage measurement." Midtronics' MDX-P300 manual adds that the key should be out of the ignition and the doors closed; its tester reports "SYSTEM NOISE" when it detects computer or ignition noise or a parasitic drain, and asks for the loads off, the doors closed and the ignition off.
2. Set the meter. Black lead in COM, red lead in the V/Ω jack, dial to DC volts. On a manual-ranging meter, Fluke says to select the 6V or 60V range depending on battery condition, and a 12-volt battery reads above 6 V, so that is the 60V range; Interstate's pro sets "0 to 20 volts".
3. Remove the surface charge if the car was just driven or charged. The next section compares how long each maker says to wait.
4. Measure at the posts. Red to positive, black to negative. Fluke: "Touch the red lead to the positive (+) battery terminal." Measure on the battery's own terminals, not on a bolt (East Penn, Midtronics) or a jump post (VARTA; Midtronics prefers the battery terminals and says a jump-post reading may be lower).

Fluke's interpretation, printed with the steps: about 12.6 volts fully charged, about 12.4 volts about 75 percent, about 12.2 volts about 50 percent, and "Below 12.0 V → Likely discharged or weak". And Fluke's rule, which East Penn's procedure also follows: "Always check voltage before load testing a battery."

The voltage is not the whole test. OPTIMA, in a 2020 post: "measuring battery voltage isn't the definitive assessment of a battery's health". It tells you whether the battery is charged enough for the next test to mean anything.

A reading below 12.0 volts. OPTIMA's no-start guide (August 2026) puts it bluntly for a roadside reading: "If the voltage is below 12.0 volts, your battery is either deeply discharged or bad and needs to be replaced." East Penn's charts put 11.8 volts at 0 percent for both flooded and AGM batteries. The honest next step is the one East Penn gives for anything below 12.4: charge it completely, then test again. A battery that passes the load test after a full charge was discharged, not dead; one that still fails after a full charge should be replaced, East Penn says.

Surface charge: why the makers disagree on how long to wait

Right after driving or charging, a battery shows a voltage higher than its real state of charge. Midtronics' MDX-P300 manual explains it in the message its tester shows: "The battery will hold a surface charge if the engine has been running or after the battery has been charged." The makers that mention it agree you should remove it; they do not agree how.

The spread is two methods: headlights to burn the surface charge off so you can test now, or OPTIMA's wait, which lets it dissipate by itself. OPTIMA's own advice after a recharge is to "check it again in another week to ten days, to make sure it is holding voltage and nothing in your vehicle is discharging your battery." A battery that keeps losing charge while parked points at a drain, and the procedure for hunting it is in car battery keeps dying.

Why it matters: the whole state-of-charge chart below spans only 0.8 volts from full to empty on East Penn's flooded scale, 12.6 down to 11.8, with 0.2 volts between rows, so a reading still lifted by surface charge can land you in the wrong row.

What the resting voltage means: the state-of-charge charts, maker by maker

The full-charge voltage is not the same across makers. East Penn's chart puts an AGM battery (12.8 V) above a flooded one (12.6 V), VARTA prints 12.8 V for a conventional starter battery, and some OPTIMA lines are higher still.

That chart is from East Penn's Form 0004 and matches its commercial poster, P162. The other makers say:

VARTA (made by Clarios), in its guide to interpreting battery tests: "A fully charged starter battery has a voltage of 12.8 Volt. If the open-circuit voltage drops below 12.4 Volt, the battery needs to be recharged."

OPTIMA, in its post on battery voltage (June 2021): "most 12-volt batteries need to measure at least about 12.6 volts before they are considered fully-charged", and "a battery could be fully-charged at only 12.5 volts or 12.8 volts". For its YELLOWTOP and BLUETOP lines: "Our YELLOWTOP and BLUETOP batteries (except for the 34M BLUETOP) are fully-charged at about 13.0-13.2 volts." And the one number it asks you to remember: "If you only remember one voltage number relating to batteries, 12.4 is the number to remember."

Interstate's video pro says "You should have a resting voltage of around 12.5 volts" and immediately adds that 12.5 volts "doesn't mean you have a good battery".

What does a resting voltage of 12.4 volts mean, according to the battery makers PulsCar read? It is the line East Penn, VARTA and OPTIMA draw. On East Penn's chart, 12.4 volts is a flooded battery at 75 percent charge, and an AGM battery between 50 and 75 percent, since East Penn puts AGM at 12.3 volts for 50 and 12.6 volts for 75. East Penn says a battery below 12.4 volts must be charged completely before a load test, because you cannot load test a discharged battery. VARTA says a starter battery below 12.4 volts needs recharging. OPTIMA says that when any lead-acid battery, AGM included, is discharged below 12.4 volts and left there, sulfation begins forming in the plates. So 12.4 volts is not a failed battery. It is a battery that needs a full charge before anyone can say whether it is healthy, and a reason to ask why it was low.

Why a full battery can still fail: charge versus health

The key sentence on this page comes from East Penn's marine and RV care brochure, written for flooded marine and RV batteries; we apply it to car batteries because it describes starting an engine: "A battery can be fully charged, but be so weak or worn out that it can no longer perform its function of starting an engine or running accessory loads." A voltmeter or a hydrometer, it says, will tell you if your battery is charged, "But these tests will not indicate if a battery can" hold a load. "Therefore, you must also perform a load test to determine the state of health of your battery."

VARTA's terms: state of charge (SOC) is what the voltmeter reads; state of health (SOH) is whether the battery can still do its job, and for a start-stop battery "the determination of the SOH takes into account the cold start capability (CCA), the residual capacity (Ah) and the charge acceptance (CA)." For an ordinary starter battery VARTA takes a softer line than East Penn: "simply the determination of the open-circuit voltage provides adequately reliable information about the state of the battery", because only the cold start current is decisive. VARTA's own warning about start-stop testers, below, is one reason this page still asks for a load or conductance test.

Cold cranking amps, the rating on the label, is defined at 0 degrees Fahrenheit (Midtronics' manuals: "The most common rating for cranking batteries at 0 ºF (–17.8 ºC)."). East Penn's FAQ explains the season: "it’s the heat that does most of the damage, but more batteries actually fail when they’re put to the test in the cold." Heat does most of the damage; cold is when more batteries fail.

So the resting voltage is a gate, not a verdict: above 12.4 volts go on to a load or conductance test; below it, charge first.

Test 3: the load test, exactly as East Penn prints it

East Penn calls the load test "the best way to determine if the battery is delivering adequate electrical performance." Its procedure, printed with the same steps and numbers in the automotive brochure (Form 0004) and the commercial poster (P162), is the most detailed load-test procedure among the documents we read. It assumes an adjustable (variable) load tester, and it opens with a rule: "First perform an open circuit voltage test, then an adjustable load test." The steps, in East Penn's words where it matters:

1. Charge first. "You can't load test a discharged battery. If the voltage is below 12.4, be sure to completely charge it before continuing."
2. Tester off, battery disconnected. "To avoid sparking and explosive gasses, be sure load tester is OFF and battery is disconnected before hook-up." Grounded cable off first, on last (East Penn's rule above). East Penn suggests a computer memory saver to keep the car's electronic memory; disconnecting has side effects on a modern car (see the AGM section).
3. Positive clamp first. Positive tester clamp to the positive terminal, then negative to negative.
4. Set the load. "Set the tester for one-half the battery‘s 0°F cold crank rating and apply the load for 15 seconds." If the CCA rating is unknown, East Penn says to use half of the minimum original-equipment CCA requirement of the vehicle.
5. Estimate the temperature, apply the load, read at 15 seconds. "Note the voltage at 15 seconds with the load on and immediately shut the load off. A reading at least equal to the value from the chart below indicates a good battery."
6. One more chance. "If the battery did not meet the required voltage and if it was not charged in Step 1, completely recharge the battery and repeat the test. If it still fails to meet requirements, replace the battery."

The pass line changes with the battery's temperature. East Penn's table for a 12-volt battery:

Three details decide whether this test tells the truth. First, the temperature is the battery's own: East Penn says to "Estimate the internal temperature of the battery to the nearest 10˚F." Second, the tester: "Make sure your variable load tester is working properly. When testing AGM batteries, make sure the tester has an AGM setting." Third, the connection: East Penn warns against clamping to bolts. "DO NOT USE Battery Bolt Extenders or Battery Bolts for testing or charging batteries," because they "can reduce your cold cranking amperage (CCA) and state of charge readings."

The marine brochure words it differently: 9.6 volts held "for 30 seconds", not 15. For a car battery, use the automotive procedure.

The fixed-load 100-amp tester: what Schumacher's manual says

A simpler kind of load tester applies one fixed load and shows the result on a dial. Schumacher's BT-100 is one: its product page describes "a 100-amp load test on 12V batteries", and its owner's manual is short and clear.

For a 12-volt battery you find the band on the meter that matches the battery's CCA rating, then: "Press and hold the load switch “on” for a maximum of 10 seconds and read the meter with the load still on." Then you read the result against the manual's table:

Two cautions from the same manual. The load heats the unit: "allow tester to cool for 5 minutes before depressing the load switch again", and "the metal housing of the unit will get hot enough to burn skin or cause property damage; carry by the handle."

East Penn scales the load to the battery (half its CCA) and reads a number at 15 seconds against a temperature table; the BT-100 applies one load and asks you to pick the CCA band on its dial. They are different tests, so do not judge a reading from one against the line of the other. East Penn says to recharge completely and repeat the test; Schumacher says a yellow or red result may mean the battery is not fully charged.

Test 4: the electronic conductance tester, and how to read its verdict

An electronic tester with two clamps and a screen, such as Midtronics' MDX-650P, which its manual calls a "Battery Conductance and Electrical System Analyzer", works by conductance rather than by drawing a large load. East Penn explains the principle in one sentence: "Conductance uses the battery’s response to a very small signal in an attempt to predict the effects of a much larger current." The words "attempt to predict" are East Penn's, and they are honest. A conductance tester does not crank anything; it infers.

Its rules for using one:

• "Conductance testing is ineffective on a discharged battery." If the battery is known to be discharged, or the tester asks for a charge, charge it completely first.
• "Be sure to distinguish between a CCA rating and a CA or MCA rating." Midtronics' MDX-650P manual defines CA and MCA at 32 degrees Fahrenheit (0°C), against 0 degrees Fahrenheit for CCA, so the three are different numbers for the same battery.
• "If the tester says to replace a battery that was tested in the vehicle, repeat the testing after removing the cables and cleaning the posts."

VARTA adds one more setting: "Set the battery tester to the correct battery type: Starter battery, gel battery, EFB or AGM battery. The device uses a different test algorithm for each battery type, so that an incorrect setting would produce an incorrect measurement value." Enter the rated cold start current with its standard (SAE, EN, DIN, JIS or IEC), printed on the label.

Midtronics' MDX-P300 manual, for a tester that handles regular, AGM and gel batteries rated from 100 to 1400 CCA, lists five verdicts and what to do with each:

Read the fourth row twice: the tester's own maker says "replace" can mean a poor connection between the cables and the battery, and East Penn agrees. Clean and retest before you buy.

VARTA's warning cuts the other way, for start-stop cars: "The indication “Battery OK” then apparently shows that the battery is in a good state of health. However, it is often obvious that the battery is reaching the end of its useful life." That applies, it says, to testers that cannot measure conductivity or have no algorithm for AGM or EFB batteries. VARTA also reads fewer start-stop events than before as the battery management system protecting a tired battery.

Extra check: the hydrometer, only on batteries you can open

VARTA says most starter batteries are now maintenance-free: "With maintenance-free batteries, there are usually no filler plugs, so that there is no possibility of measurement with a hydrometer." If yours has removable caps, East Penn's marine and RV brochure (written for flooded batteries) gives the procedure and the line: take a hydrometer reading from one cell, using a different cell each time, and "If the reading is above 1.225 specific gravity, the battery is at least 75% charged and should be OK."

The reading must be corrected for temperature, to 80 degrees Fahrenheit (27°C). East Penn: for each 10 degrees above 80°F add four points, and for each 10 degrees below, subtract four. Its own worked example: "at 90ºF (32ºC) a hydrometer reading of 1.250 would be corrected to 1.254."

Two rules from the same page. "Never add acid, only water to a battery." And for this test above all: "FLUSH EYES IMMEDIATELY WITH LARGE QUANTITIES OF COOL WATER. GET MEDICAL HELP FAST." Glasses (East Penn), protective gloves (Clarios) and, as Schumacher's manual says, someone nearby for help.

The hydrometer answers the same question as the voltmeter, how charged the battery is. Like the voltmeter, it cannot tell you whether the battery can deliver cranking current. East Penn's marine brochure puts the voltmeter and hydrometer side by side and then says both need a load test to judge health.

Test 5: cranking voltage, and when the starter is the problem

With a multimeter on the battery posts, the engine is cranked and you read the voltage; Fluke's guide says to "Use MIN/MAX to catch voltage drops when starting the engine."

The pass line in Midtronics' PBT-300 manual: "The cranking voltage is greater than 9.6 V. The starting system is OK." The manual says the battery must be good and fully charged for this test. Below that: "The cranking voltage is less than 9.6 V, which indicates a starting system problem. Check the connections, wiring, and starter." Midtronics' MDX-P300 separates the cases by the battery's state: "CRANKING VOLTAGE LOW" means the starter voltage was low with a fully charged battery, and "CHARGE BATTERY" means the voltage was low because the battery was discharged, so charge and repeat.

Schumacher's BT-100 manual ties the cranking line to the load-test result: its table runs from a minimum cranking voltage of 7.7 volts after a 10.2-volt load reading up to 10.6 volts after 11.4. It is a starter current-draw test that cannot be run on a battery that tested weak or bad, and it asks for a warm engine, the ignition disabled if possible, and a crank of 2 to 5 seconds.

Voltage drop across the cables. Fluke: "Slow cranking or intermittent starting often points to excessive resistance in a vehicle’s starter circuit." Its starter circuit voltage drop guide measures across the cable itself while cranking: red probe on the battery positive post, black on the starter's battery stud, then the ground side from the engine block to the negative post. Its line: "Acceptable total drop: 0.2 to 0.5 V", about 0.3 volts on the positive side and 0.2 or less on the ground side; "Anything higher suggests corrosion, loose terminals, or damaged cables." Its safety list for this test: park or neutral with the parking brake on, "Keep hands and leads clear of moving parts and avoid shorting terminals with probes", and "Disable ignition or fuel (e.g., remove the fuel-pump relay) so the engine cannot start during the test."

Clicks. OPTIMA's no-start guide reads the sound: "If you hear a rapid clicking noise or your engine sounds like it's trying to start, but sounds like it's struggling and cranking slowly, your battery could be deeply discharged." One solid click and no crank is different: "The starter solenoid could be engaging, but the starter motor itself isn't turning." The paths after that are in car clicking when starting, car won't start, no click and starter replacement cost.

Test 6: is it the alternator? Charging voltage, with the loads on

Once the battery is charged and tested, the alternator question is a voltage reading with the engine running. Denso, an alternator maker, asks for a fully charged battery at 12.6 volts before any charging-system test. Run the engine in the open, not a closed garage: OSHA's carbon monoxide fact sheet says "Large amounts of CO can overcome you in minutes without warning". Parking brake on, park or neutral (Fluke's starter-circuit guide), someone nearby (Schumacher), and hands, leads and clothing clear of belts and pulleys. Then the makers' numbers:

The bands overlap more than they differ: where makers print a range, it starts between 13 and 13.7 volts and tops out between 14.5 and 15.5 (Interstate gives only a floor, the MDX-P300 only a typical regulator limit). Where they differ most is the top end, and Midtronics gives the reason in the same manual: the regulator sets the voltage from battery voltage, under-hood temperature and the loads in use, so "Refer to the manufacturer specifications for the correct limit, which may vary by vehicle type." Your car's service information beats every table here.

How can you tell whether it is the battery or the alternator, using the numbers PulsCar collected from Denso, Midtronics and Fluke? Charge the battery fully first; Denso asks for 12.6 volts before any charging test. Then measure at the battery with the engine running at about 2,000 rpm, and again with the headlights, blower and defrosters on. Denso's troubleshooting page wants 13 to 15 volts in both cases; its diagnosis page says a voltmeter gauge reading below 13.2 volts could indicate a failed alternator or bad wiring, and above 14.5 volts a failed regulator, a bad battery or poor connections. Midtronics' PBT-300 passes 13.3 to 15.5 volts and Fluke's guide calls 13.7 to 14.7 volts a good alternator. If the charged battery then fails a load test, the battery is the problem. If it passes but the running voltage is low, the alternator, its belt or its cables are next.

Loads on, not just idle. Schumacher's BT-100 test runs the engine at 1,200 to 1,500 rpm and checks that the reading stays in the green band "With the headlights and blower motor on high". OPTIMA explains what dimming headlights mean: "it means your vehicle's charging system is not able to meet the electrical demand at that moment and the battery is trying to cover the gap."

The belt before the alternator. Midtronics: "Before starting the test, inspect the alternator drive belt." A belt that is glazed, worn or loose, the manual says, will keep the engine from reaching the rpm the test needs. For a charging voltage of none, its manual's order is belts, then connections, then the alternator; for low, belts, then connections. Denso adds fuses and fusible links.

If the charging test points at the alternator, what a replacement costs is in alternator replacement cost.

Ripple and voltage drop: two tests that catch a half-failed alternator or a bad cable

An alternator makes alternating current and its diodes convert it to direct current; Fluke says a single open diode "lets AC leak into the system, increasing ripple." Fluke's ripple voltage guide says an abnormal rise in ripple "often signals a failed rectifier diode or damaged stator winding — faults that bring flickering lights, battery-drain complaints, or erratic engine behavior long before a total charging failure."

The test needs a meter with a low AC range: loads on (high beams, blower, rear defogger), engine idling for about 30 seconds, meter set to AC volts, red on the alternator output terminal or the battery positive post, black on ground. Fluke's lines:

The ripple guide adds a caution before blaming the alternator: "High ripple can stem from poor alternator or battery grounds. Verify clean, tight ground straps before replacing the alternator." Keep hands, leads and clothing clear of belts and pulleys; Fluke says the battery posts are "electrically equivalent to studs and safer for the operator."

Charging-circuit voltage drop. Denso measures the cables the same way Fluke measures the starter cables. With the engine at 2,000 rpm and the lights, blower and radio on: meter positive on the alternator output stud and negative on the battery positive post should read "less than .2 volts"; meter negative on the alternator case and positive on the battery negative post, ".2 volts or less". "High voltage drop indicates poor connections or damaged cables."

Battery or charge light on while driving: stop safely first

A battery or charge warning that comes on while you are driving is a road problem first. PulsCar's study advises: "Turn off what you can, head for the nearest safe place to stop, do not shut the engine off until you are there". Test 6 is what you do once you are stopped and safe. The study found cars that stopped at speed, in traffic or on a highway, and notes that "Some warnings are not about the battery at all". The broader guide to what each dashboard light means is in dashboard warning lights.

Denso's diagnosis page describes the charge lamp's normal behavior: with the ignition on and the engine not running "The charge lamp should illuminate", and with the engine running "The charge lamp should illuminate briefly then turn OFF." A weak battery during high current draw, a low idle (dimly) or corroded wiring can also light it, Denso says.

PulsCar's study of battery and alternator complaints found that 17.2 percent of 10,776 federal complaints about a failing 12-volt battery or charging system describe the car dying, stalling or losing power while being driven; 25.2 percent when the owner names the alternator. These are shares of what owners reported, not failure rates (method and data).

AGM, EFB and start-stop batteries: the test changes

Three things change when the battery is not a plain flooded one.

The full-charge voltage is higher. East Penn's chart puts a full AGM battery at 12.8 volts against 12.6 for flooded, and 75 percent at 12.6 against 12.4. A reading of 12.6 volts is a full flooded battery and a three-quarters AGM. Check the label before you read the chart.

The tester must know. East Penn: "When testing AGM batteries, make sure the tester has an AGM setting." VARTA says the battery type setting selects a different test algorithm, and that for AGM and EFB batteries "it is essential to use modern test devices which are adapted for the new battery technologies." A tester without an AGM or EFB mode can pass a tired start-stop battery, which is VARTA's warning quoted above.

Charging is different. East Penn: "Warning: Gel and AGM (Absorbed Glass Mat) batteries require a voltage-limited charger. Charging a Gel or AGM battery on a typical shop charger that exceeds 15.4 volts – even one time – may greatly shorten its life." For the rare AGM that needs charging outside the vehicle's charging system, on a charger, it gives 13.8 to 14.8 volts at 77°F (25°C), not to exceed 30 amps. If you charge before retesting, use a voltage-limited charger that is AGM compatible.

Replacement is not free choice. VARTA's guide to EFB and AGM: "Cars with an extended automatic start-stop system are equipped with an AGM battery. Only an AGM battery may be used as a replacement." VARTA adds that the wrong battery can cause "deficiencies in the automatic start-stop system, failures of comfort functions and reduced service life of the battery." The type mistake and what it costs are covered in car battery replacement cost.

Disconnecting has side effects. East Penn's load test disconnects the battery; its FAQ explains why that is no longer trivial. For most modern vehicles, East Penn says, once battery power is disconnected the engine computer's learned data, such as misfire and oxygen-sensor records, is cleared and must be relearned, which East Penn says can mean "failing an emission’s test (depending on the state of inspection)". "Several BMW, Audi, and Lexus vehicle models require the computer system to be reset after battery replacement." If you have an emissions inspection coming, read how to reset a check engine light before you disconnect anything; a conductance test or a voltmeter reading on the connected battery disconnects nothing.

Hybrids and electric cars carry a 12-volt battery too; PulsCar's complaint study counts 12-volt failure complaints from both. We did not read a hybrid or EV maker's test procedure, so your owner's manual is the source; costs are in car battery replacement cost.

Charge before you retest: why driving around does not count

East Penn's load and conductance tests and Midtronics' "CHARGE & RETEST" verdict all require a fully charged battery before the next test, and OPTIMA says the car's own alternator is not the tool for it; Interstate says driving would take hours at highway speeds and strain the alternator.

OPTIMA: "Your vehicle's charging system is designed to MAINTAIN a battery near a full state of charge, not RECOVER a deeply discharged battery." It warns that trying can cause "a cycle of dead batteries and jump starts" until either the battery or the charging system needs replacing. Interstate, on the same point: "Jump leads do not charge a battery."

How long a proper charge takes depends on the battery and the charger. OPTIMA's worked example: "If you have a 50Ah battery and your battery charger is charging at 10 amps, it could take more than five hours to fully recharge your battery." East Penn prints a charging-time chart by state of charge and charger rate, and a rule its commercial poster P162 gives for maintenance-free flooded batteries: the maximum charge rate in amps "should be no more than 1/3 of the battery’s reserve capacity minute rating", and "If the terminal voltage exceeds 16.0 volts while charging, reduce the charge rate." Both are flooded-battery rules; for gel and AGM its limit is the 15.4 volts above.

Clarios's safety sheet has the one rule for where: "Never recharge batteries in an unventilated, enclosed space." Charger off before clamps go on or come off, never on a frozen battery (East Penn), and follow the charger maker's connection order.

Then remove the surface charge before you test (the table above compares the makers' methods). After a full charge, East Penn calls a reading at least equal to its chart a good battery; one that still fails is, in East Penn's and Midtronics' words, the one to replace.

Free battery tests at parts stores: what their own filings say

You do not have to own a load tester. AutoZone, O'Reilly and Advance Auto Parts each describe battery testing in their annual reports to the Securities and Exchange Commission, and O'Reilly also in its newsroom; the three chains' battery-testing pages answered our requests with "Access Denied" on 30 September 2026.

Interstate offers one too, at its "Interstate All Battery Center locations", and OPTIMA says "Most quality battery professionals can load test your battery and many will do it for free."

None of the filings says which tester the store uses or whether it charges a low battery before testing it (AutoZone lists battery charging as a separate free service), and Advance's "except where prohibited" does not say where. Ask at the counter, because a conductance test on a discharged battery is the case East Penn calls "ineffective". Ask for the measured voltage and the rating entered, and if the verdict is "charge and retest", do that before buying.

Jump-starting to get to a test: East Penn's order

A jump gets a flat car to a charger or a test. East Penn's brochure begins its instructions with cautions, including: "PRIOR TO JUMP-STARTING VEHICLE, REFER TO THE VEHICLE OWNER’S MANUAL" (some cars have remote jump terminals), and "To avoid a battery explosion, never attempt to jump start a frozen battery." Its order:

1. The vehicles are not touching and both ignition switches are off.
2. Positive clamp to the positive terminal of the stranded vehicle.
3. The other positive clamp to the positive terminal of the servicing vehicle.
4. Negative clamp to the negative terminal of the servicing vehicle.
5. The final negative clamp to the engine block or frame of the stranded vehicle (or its negative remote terminal, if it has one).
6. "Start the servicing vehicle and allow it to run for at least 30 seconds. Then, start the stranded vehicle and remove the booster cable in REVERSE order of the above connections."

Interstate's jump-start page connects the positive clamps starting with the good car rather than the dead one, but agrees on the part that prevents sparks at the battery: the last black clamp goes "on a metal component like a bracket, not the dead battery." It adds: jewelry off, clamps kept apart, and a look at the battery first. Its step after a successful jump is a battery test, and OPTIMA's is a full recharge: "you still need to fully recharge your car battery with a battery charger and you need to do it as soon as possible." The jump did not charge it.

The traps: three ways a battery test misleads

Trap one: testing a battery straight after a drive. The surface charge lifts the resting voltage, so the reading can look better than the battery's real state of charge; Midtronics: the battery holds one "if the engine has been running". Midtronics, Fluke (as an optional step), Interstate and VARTA (before a conductance test) describe removing it first, with headlights for 30 seconds to two minutes where a time is given, or, OPTIMA's version, to measure after the car has sat 12 to 24 hours. A reading taken in the parking lot right after the drive to the store is not a resting voltage.

Trap two: judging a discharged battery on a load or conductance tester. East Penn is flat about it: "You can't load test a discharged battery", and "Conductance testing is ineffective on a discharged battery." A low battery can fail either test and still be healthy once charged. Midtronics builds the step into its verdicts: "CHARGE & RETEST" means exactly that, and a second CHARGE & RETEST after a full charge means replace (its separate REPLACE verdict can be a poor connection, Trap three). The trap is buying a battery on the first result.

Trap three: testing through a bad connection. A corroded clamp, a bolt instead of the lead post, or a remote jump post instead of the battery puts resistance in the test. Midtronics says its "REPLACE BATTERY" result "may also mean a poor connection"; East Penn says to retest after removing the cables and cleaning the posts; VARTA says the tester goes on the battery, "and not the starting aid contacts". The same trap has a charging-system version: a low running voltage at the battery can be a cable, and Denso lists a voltage-drop test of the charging cables among its system tests.

Three composite situations, decoded

These are composite scenarios, not individual customers. Each is built only from facts already on this page.

The battery that "passed" in the parking lot. An owner has a slow crank on a cold morning, drives to a store, and a tester says the battery is fine. Was the surface charge removed after the drive (Trap one)? The plan: at home, after the car has sat for 12 to 24 hours (OPTIMA's wait), measure the resting voltage, charge fully if it is below 12.4 volts, and have a load or conductance test done on the charged battery with the right type (flooded, AGM or EFB) and the rating from the label entered.

The new battery that went flat again. A battery was replaced, and the car will not start again. The page's order says: charge it fully, confirm it at rest, then run Test 6. A charged battery with a running voltage below the 13.2 volts Denso flags on a voltmeter gauge, or a PBT-300 highest average outside 13.3 to 15.5, points at the belt, the cables or the alternator. If the charging voltage is normal, the next suspect is a drain, and the procedure is in car battery keeps dying.

The "replace" verdict on an old-looking battery. A conductance tester at the counter says replace. The battery terminals are crusted. Before paying, note what a tester maker and a battery maker both say: Midtronics, that the result "may also mean a poor connection between the battery cables and the battery"; East Penn, to repeat the test after removing the cables and cleaning the posts. If a clean, charged battery still fails, it is done, and the type on the label decides what goes in.

Where a free sound check fits, and where it does not

This is the part of the page about our own product, so check it against our site. The homepage terms of the sound check are "No payment, no card, no subscription." If a clip cannot support a confident verdict, the homepage says, "your report honestly says “inconclusive”" rather than guessing.

Can PulsCar's free sound check test a car battery? Not in the sense this page means, and it is worth being exact about why. The sound check listens to a 30-second phone recording and emails a report in about 10 minutes. A battery test is an electrical measurement: resting voltage, voltage under a load of half the cold cranking amp rating, conductance, cranking voltage and charging voltage, each read at the battery posts with a meter or tester, as East Penn, Midtronics and Fluke describe. A microphone cannot read any of those numbers. PulsCar's homepage lists a starting and charging category, including clicking when starting, so a recording can suggest where to look, but it is not a measurement and does not replace Tests 2 to 6. It fits best after the battery passes: an engine that runs but sounds wrong, such as a squealing belt, is what PulsCar was built to listen to.

So the order stands: meter, then a load or conductance test on a charged battery, then the charging voltage. Sound versus scanner makes the case for when each tool fits.

What we looked for and did not find

Naming the gaps is part of the method. Each of these was an actual request or search on 30 September 2026.

Odyssey's technical manual. Odyssey's battery site (odysseybattery.com and its eu. subdomain) answered our requests for its technical manual and owner's manual with HTTP 403. We did not use copies hosted elsewhere; nothing here comes from Odyssey.

The parts stores' own battery pages. autozone.com, oreillyauto.com and shop.advanceautoparts.com each answered with "Access Denied". We did not work around the blocks. What the stores offer is quoted from their annual reports and, for O'Reilly, its newsroom, and nothing here describes which tester a store uses or how it runs the test.

OPTIMA's charging and support FAQ. Those pages showed no text when we opened them; we quote OPTIMA's dated blog posts instead.

A Battery Council International test procedure. BCI is Battery Council International; Midtronics' manual tells users to "Follow all BCI (Battery Council International) safety recommendations." We did not find a BCI testing procedure we could read, so none is quoted.

One agreed full-charge voltage. The makers we read do not agree on one: East Penn prints 12.6 volts for flooded and 12.8 for AGM, VARTA prints 12.8 for a starter battery, OPTIMA says about 12.6 and as low as 12.5 or as high as 12.8, and about 13.0 to 13.2 for its YELLOWTOP and BLUETOP lines (except the 34M BLUETOP). The page shows all of them rather than inventing an average.

Your action plan: tonight, tomorrow morning, and this week

Tonight (a multimeter):

1. Look at the battery. Cracked or broken case, or a rotten-egg smell: stop (Schumacher: do not use the tester on a damaged battery; Interstate: keep away from one that smells), and get it to a professional. Bulging: OPTIMA says replacement should be a priority.
2. Glasses on, jewelry off. Clean crusted terminals first, neutralizing with baking soda and water as East Penn describes.
3. If you just drove, headlights on for about two minutes, then off (Interstate); if the car has sat for hours, skip this. Measure at the posts and write down the number.
4. If the resting reading was about 12.6 volts or more, start the engine and measure again at the battery, then with headlights, blower and defrosters on, and compare with Test 6, keeping hands and leads clear of belts and pulleys (Fluke). If it was lower (or below about 12.8 volts for an AGM, by East Penn's chart), charge first: Denso asks for a fully charged battery before any charging test.

Tomorrow morning, before you start the car: 5. Measure the resting voltage again and compare it with the East Penn chart for your type (flooded or AGM, from the label). If it is full, set the meter to MIN/MAX on the posts while a helper cranks: the PBT-300 line is above 9.6 volts; below that on a charged battery, Midtronics says check the connections, wiring and starter. 6. Below 12.4 volts (the charge line East Penn and VARTA print): charge it fully with a charger that suits the type, then wait and measure again. Above: go to step 7.

This week: 7. Get a load test or a conductance test on the charged battery, yourself or free at a parts store. Ask for the measured voltage and the rating entered, and for a retest after cleaning the terminals if the verdict is "replace". 8. If the battery passes but the running voltage was out of band, check the belt, the cables (Denso's 0.2-volt drop test) and then the alternator; alternator replacement cost has the prices. 9. Everything passes but the battery still goes flat: the drain test in car battery keeps dying. It fails after a full charge: replace it with the type the car was built for (car battery replacement cost).

Sources and documents

Every quotation above comes from one of these documents, each opened and saved as text on 30 September 2026. Pages that refused to open are listed as such and are not quoted.

1. East Penn Manufacturing (Deka), Basic Rules for Battery Care & Maintenance, Form No. 0004, dated 02/26 — safety, inspection, state-of-charge chart, load and conductance tests, charging, jump-starting
2. East Penn, Commercial Battery Care & Maintenance poster P162, Rev. 01/22, and Automotive poster P121 — the same procedure and chart, read to cross-check
3. East Penn, Marine/RV Battery Care & Maintenance brochure 0273 — hydrometer, charge versus health, the 30-second marine line
4. East Penn, Frequently Asked Questions — heat and cold, system resets after replacement
5. VARTA Automotive (Clarios), Testing the battery, step by step, Interpretation of battery tests and EFB and AGM differences, published April and May 2024
6. OPTIMA Batteries (Clarios), What are the Signs a Battery is Dying? (24 September 2026), Why Won't My Car Start? (28 August 2026), Basic Information on Battery Voltage (30 June 2021), Time To Check Your Battery Voltage! (15 September 2020)
7. Interstate Batteries, How to Test a Car Battery with a Multimeter (video transcript) and The Official Way to Jump-Start a Car Battery
8. Clarios, Safety Data Sheet, Lead Acid Battery, revised 12/13/2023, linked from its safety sheets page
9. Midtronics, instruction manuals for the MDX-P300, PBT-300 and MDX-650P SOH — verdicts, surface charge, cranking and charging lines, rating definitions
10. Schumacher Electric, BT-100 owner's manual, linked from its product page
11. Fluke, How to Measure Battery Voltage with a Multimeter, alternator ripple voltage and starter circuit voltage drop, updated February 2026
12. DENSO Auto Parts, Charging System Diagnosis and Charging System Troubleshooting
13. OSHA, 29 CFR 1926.441, batteries and battery charging, and 29 CFR 1910.178 paragraph (g) — workplace rules, quoted as context; OSHA Carbon Monoxide Poisoning fact sheet, FS-3522, 04/2012
14. Annual reports on Form 10-K: AutoZone (year ended 30 August 2025), O'Reilly (2025), Advance Auto Parts (year ended 3 January 2026); O'Reilly newsroom, 1 December 2025
15. PulsCar, battery and alternator complaint study, its method file, and the homepage for the sound check terms
16. Could not read on 30 September 2026, not quoted: odysseybattery.com technical and owner's manuals (HTTP 403); the battery-testing pages of autozone.com, oreillyauto.com and shop.advanceautoparts.com ("Access Denied"); optimabatteries.com charging and support FAQ (no text in the page as delivered)

How this page was built: every procedure, pass line and safety instruction is quoted or summarized from a document by a battery maker (East Penn, VARTA, OPTIMA, Interstate, Clarios), a tester or meter maker (Midtronics, Schumacher, Fluke), an alternator maker (Denso) or OSHA, each saved as text on 30 September 2026. The one roadside instruction, for a charge warning while driving, is from PulsCar's own study, and the action plan's order is our arrangement of those documents. Parts-store services come from the companies' SEC annual reports and O'Reilly's newsroom, because their battery-testing pages refused our requests; we did not work around those blocks, or Odyssey's. We ran no tests ourselves. Complaint figures are PulsCar's own published study, cited with its limits. The three scenarios are composite and contain no invented names, amounts or dates.

Prices verified: September 2026 (no prices are quoted on this page) · Technical documents verified: September 2026 · Reviewed quarterly

Tested your battery and got a result this page does not explain, or found a maker document that contradicts a number here? Email [email protected] with the document or the reading. A documented counter-example changes this page faster than anything we could write.