P0456 Code: The Smallest EVAP Leak, and What 133 Complaints Say It Really Was

P0456 is a measurement, and it is a smaller measurement than most people realize. Your car sealed the fuel vapor system, watched the pressure, and decided that what leaked away was at least as much as would escape through a hole half a millimetre across.

That is the whole claim. It does not name a part, it does not say where, and it does not say it is one hole.

The cap is still the right first thing to check, because it costs nothing and manufacturers genuinely put it first. But by the time a P0456 reaches a repair shop it is usually past that. In all 133 federal complaints naming this code that we read for this guide, only twelve owners mention the cap at all — and in exactly one does replacing it appear to be the answer.

By the end of this guide you will know which of five routes your P0456 belongs to, what proves it, what each one costs, and which VINs get the repair free. Claims marked 🟢 rest on a document in Sources, opened and read; ⚪ marks estimates and owner reports.

I built Pulscar — an AI tool that reads car problems from a 30-second recording — after a $380 "full diagnostic" in Phuket in 2025 led to roughly $6,000 over six months across several shops, and the sound never went away (the story is on the About page).

First: is it really P0456? The EVAP codes side by side

Write down every code stored with it. The neighbors change the route more than the code does.

The codes are a ladder of hole sizes, not a ladder of seriousness. A P0455 often means something is open — a cap left off, a line disconnected — and is frequently the cheapest of the three to fix. P0456 is the one that hides.

P1450 changes everything on a Ford. Twenty-one of our 133 complaints carry P0456 and P1450 together, and nearly all of them describe the same story: the car stalls or refuses to start after a fill-up. That is not an emissions complaint, and it has a recall attached.

A fuel-tank-pressure-sensor code is not a leak code. P0451 and P0452 say the instrument is wrong, not that the system is leaking. Fix the instrument before believing the measurement.

What 0.020 inch actually means, in the regulation's own words

This page's one advantage over every generic code definition is that the threshold is written down, and we read it.

California's OBD II regulation, 13 CCR 1968.2, section (e)(4.2.2)(C), requires the system to report a malfunction when "The complete evaporative system contains a leak or leaks that cumulatively are greater than or equal to a leak caused by a 0.020 inch diameter orifice." Subsection (B) sets the same requirement at 0.040 inch, which is what P0455 reports.

Three words in that sentence decide how you should diagnose the car.

"0.020 inch" is 0.508 mm. You are not going to find this by looking. Soot streaks, wet patches and visible splits belong to bigger leaks; this one is found with smoke, pressure or the car's own test, and nothing else.

"Cumulatively" means the threshold is the sum of everything leaking. Two seepy O-rings and a slightly tired cap seal can add up to one failing test while no single one of them would. That is why owners replace a part, see the light return, and conclude the part was fine — sometimes it was, and the remainder was still over the line.

"Orifice" is defined in section (e)(4.2.1) as "an O'Keefe Controls Co. precision metal 'Type B' orifice with NPT connections with a diameter of the specified dimension (e.g., part number B-20-SS for a stainless steel 0.020 inch diameter orifice)." The threshold is a purchasable part with a part number. That is what a proper leak tester is calibrated against, and it is a fair thing to ask a shop about.

Why this code comes and goes, and why that is written into the rules

Owners describe the same maddening pattern over and over: the light appears, disappears, and returns weeks later. There is a published reason.

The same regulation sets how often each monitor has to run. Section (d)(3.2.1)(B) gives a minimum acceptable in-use monitor performance ratio of 0.260 for monitors designed to detect a 0.020 inch leak and 0.520 for purge flow and 0.040 inch leak detection. In plain terms: the smallest-leak test is only required to complete half as often as the large-leak test.

That single number explains most of the frustration around this code.

A cleared code takes longer to come back. Not because the leak healed, but because the test that finds it runs least often of all the evaporative tests.

Readiness before an inspection takes longer to set. If you are waiting for monitors after a repair or a battery disconnect, this is usually the last one to complete.

"I cleared it and it was fine for two weeks" proves nothing. Two weeks is well inside the window this monitor is allowed to take.

You can read the result as a number. Section (g)(4.5.4)(F) requires that a system with separate 0.040-inch and 0.020-inch diagnostics "shall separately report 0.040 inch and 0.020 inch test results." Any scan tool that shows Mode 6 can show you the 0.020-inch test result against its limit — a number, not an opinion.

The 5-minute driveway protocol

1. Read every code and write them all down before anything is cleared. A $20 to $40 Bluetooth reader gives you stored, pending and history codes plus the freeze frame. P1450, P0455 or a misfire beside P0456 changes the route.
2. Open the filler, look at the cap seal, and refit it properly. Wipe the rubber seal and the metal lip it meets. Turn until it clicks, more than once on some trucks. Look for a cracked, hardened or missing seal. This is the free step, and it is the one manufacturers put first.
3. On a capless filler, look into the flap instead. There is no cap to blame. Ford, Honda and Acura owners in our set found leaks at the flap and the filler assembly, not at a cap they do not have.
4. Smell around the rear of the car with a full tank. A strong fuel smell is not a 0.020-inch leak. It is a bigger one, and it moves you to Route 3 and to a sooner appointment.
5. Note what happens right after a fill-up. Hard start, no start or a stall in the first few miles after refuelling is the canister-purge-valve pattern: Route 2 and Route 5, not the cap.
6. Run the VIN through NHTSA's recall lookup. On some Ford and Volkswagen cars this step ends the story at $0 — and on the Ford cars the recall is about a stall, not a leak.

Five minutes, a phone and a cheap reader, and you have done the part of the diagnosis that nobody charges for. What a diagnostic should cost covers the rest.

The price ladder: read it from the bottom up

Read it from the bottom up. On this code the distance between the top of the ladder and the bottom is almost entirely a question of whether anybody located the leak before buying a part.

Your number, by what you drive

Two rows in that table are worth reading twice. On the Focus the code sits on a stall recall, so the urgency comes from the stall and not the leak. On the Nissan filler tube bulletin the manufacturer published a remedy in February 2026 and withdrew it in May — which is a reminder that a document your dealer quoted six months ago may no longer be the current answer.

The five routes, cheapest proof first

Route 1 · The cap and the filler — the free check, and what the file says about it

The tell. The light appeared within a few days of a fill-up, and there is no smell, no stall and no other code.

What it actually is. The cap is the only part of a sealed system a driver opens on purpose, so it is the most likely thing to be left unsealed. Mazda says it four times over, in bulletins covering 2007 to 2019 models: the concern "is most likely the result of the customer not getting the fuel cap completely closed after fueling the vehicle." Nissan and Infiniti open their EVAP leak guidance for every model year from 2012 onward with the same idea — "a loose fuel fill cap may be the cause" — and then tell the technician to confirm it with a leak test rather than assume.

The cap is also more than a cap. Hyundai replaced the cap on 2017 Santa Fe Sport cars where "the fuel filler cap seal may not provide proper sealing." Toyota traced a 2014-2016 Corolla cap that gives "a 'No Click' and/or 'No Spring Return' condition" to a filler neck sitting out of position, so the part at fault was the neck. And on capless Hondas and Acuras, the tech line article is blunt: "don't replace the fuel fill pipe! Instead, replace the mis-fuel inhibit device and the main flap unit."

The cheap cause that gets skipped. The seal itself, and the lip it seals against. Wipe both. A ring of grit on the filler lip is enough, and it costs nothing to remove.

Fix it yourself — refit it properly, then wait properly. Cold engine, ignition off. Remove the cap, wipe the rubber seal and the metal lip, check the seal for cracks or hardening, refit and turn until it clicks — on some trucks that is several turns. Then drive normally for a week before you judge it, because the monitor that has to clear this code is the least frequent one on the car. The honest limit: if the seal is missing, torn or the cap will not click, buy the manufacturer's part number rather than a universal cap.

Where this does not apply. If you have a capless filler there is nothing to tighten, and if the code came with P1450 or a stall the cap is not the story.

At the shop — say this: "Before any parts, confirm or rule out the cap and the filler seal with a leak test, and tell me which one you found. If the leak is elsewhere, show me where."

Route 2 · A valve that leaks while it is closed — purge, vent and NVLD

The tell. No smell, no visible damage, and often a rough idle, a hard start or a stall in the first minutes after a fill-up. On the reader: P1450, P0451 or P144A beside P0456.

What it actually is. The evaporative system is sealed by valves, and a valve that does not seal completely fails a 0.020-inch test without a hole existing anywhere. This is where the evidence points hardest. In our reading of all 133 complaints, a purge, vent or EVAP valve is the named cause in 38 — 28.6%, the largest single group. In the manufacturer documents it is the largest group too: of the 44 distinct problems we sorted out of the 77 non-Stellantis bulletins, 10 are purge, vent or NVLD valves.

The documents also tell you which valve, and sometimes which part of it. Ford's 17-0016 blames "an intermittent sticking evaporative emission canister purge valve" on the Focus RS. Ford's 19-2207 traces the same codes on 2013-2017 hybrids to "the sealing capability of the canister purge valve" and pairs a software update with the valve. Kia's TSBFUE040 replaces a canister close valve "stuck open" on 2012-2017 Rios. And Nissan's NTB17-082, with its Infiniti twin, is the most useful document on this page: it is titled "DTC P0456 CAUSED BY EVAP LEAK AT VENT CONTROL VALVE O-RING," it requires the technician to confirm first that "DTC P0456 is not caused by a loose fuel fill cap," and then instructs: "DO NOT REPLACE the vent control valve. Replace the vent control valve O-ring seal."

The cheap cause that gets skipped. The test that decides. Ford's SSM 50306, covering 21 models from the 2018 EcoSport to the 2022 Maverick, says Ford "has tested returned EVAP canister purge valves and found them to be fully functional" and asks for a full system leak test and a purge valve leak test "prior to replacing the canister purge valve or other EVAP components." A companion message, SSM 46525, adds a check for liquid fuel in the vapor lines, which "should not be present," before any valve is replaced.

Fix it yourself — one free test on a stored purge valve. On many cars the purge valve is at the top of the engine on a vapor line you can unclip by hand. With the engine off, disconnect the line and blow gently through the valve: a closed purge valve should not pass air. One owner in our set documented that exact check on a 2017 Fiesta and named the part number that failed it. The honest limit: this only catches a valve stuck open. A valve that seeps only under vacuum, an NVLD sensor and a vent valve at the canister all need a proper leak test.

Where this does not apply. If you can smell fuel, the leak is bigger than a valve seat: Route 3.

At the shop — say this: "Has the purge valve been leak-tested rather than swapped? Ford's own message says returned valves tested functional. And if the leak is at the vent valve, is it the valve or the O-ring?"

Can a sound check tell you anything about an evaporative leak? Not about the leak itself, and this page will not pretend otherwise: a 0.020-inch hole is silent. What it can hear is the company this code keeps. Forty of our 133 complaints describe a stall, a no-start or a rough idle alongside P0456, and almost all of them are canister-purge-valve cars where the valve is upsetting the running engine, not just the emissions test. Film 30 seconds of the engine running, parked, and Pulscar's free check returns a likely cause, a severity and a repair-cost estimate, or "inconclusive" — it does not decide whether the car is safe to drive. Free, no card, about ten minutes, and the report arrives by email. If the idle is clean, that points you back at the sealed system rather than the engine.

Route 3 · A crack you can smell — tank, filler tube, vent line and pump seal

The tell. A smell of gasoline, worst with a full tank or right after filling. Rust or road-salt exposure. Slow filling, or the nozzle clicking off early. Recent work near the tank.

What it actually is. Everything downstream of the filler is plastic, rubber and steel that ages. In our complaint reading 18 of 133 stories end at the tank, the filler tube, a vent tube or a hose — cracked rollover valves on two Fords, a fitting cracked on top of a Mustang tank, a Jeep Wrangler vent tube that cracked twice, a Mitsubishi filler neck corroded through in Pennsylvania, a Suzuki fill pipe with rust holes in the vent pipe. Manufacturers document the same class of failure with smaller remedies: Ford ties P0456 on 2022-2024 Mavericks to "a chaffing or rubbing condition between the fuel vapor canister line and the air vent hose", and fixed a kinked fresh air vent line on Transits by attaching a zip tie.

Two more locations are worth knowing because they are easy to miss. Kia's PS567 covers "DTC P0456/P0442 WITH FUEL PUMP SEAL DAMAGE" on 2017 Cadenza and Optima Hybrid cars, and says to look there only "when the vapor lines, fuel cap, and all other corresponding circuits are found to be within specifications." Kia's PS322 and Hyundai's canister bulletins both point at cracks where the charcoal canister bolts on — Kia asking technicians to "closely inspect the mounting holes of the charcoal canister for visible cracks" after the lines and cap check out.

The cheap cause that gets skipped. A clamp, a zip tie or a rerouted hose. Two of the five line-and-hose problems in the bulletin set are fixed with hardware that costs less than the diagnostic fee, and both were found by looking rather than by testing parts. The gas smell guide covers a smell with no code.

At the shop — say this: "There is a fuel smell with a full tank. Smoke-test the whole system including the filler tube and the top of the tank, and photograph the leak before you quote."

Route 4 · Nothing is leaking — software, calibration and a monitor that is wrong

The tell. A smoke test found nothing, a new cap changed nothing, the engine runs perfectly, and the light keeps returning.

What it actually is. Sometimes the measurement is wrong rather than the system. Audi wrote the cleanest example we found, in a 2015 Q7 bulletin whose entire summary reads: "The MIL is on due to DTC P0456 being stored in the ECM. There is no leak in the EVAP system." A companion bulletin for 2014-2015 Q7s adds that "Drivability is not affected." Ford says the same thing on newer trucks in almost the same words: on 2021-2023 Broncos with the 2.7L and 2022-2024 Explorers, the light "may be due to the software level of the PCM and has no effect on vehicle operation."

Calibration can be wrong in a more specific way. Ford's 21-2020 traces P0456 on 2020 Escape 2.0L front-wheel-drive cars built on or before 20 March 2020 to "the calibration in the powertrain control module (PCM) being configured for an L-tank instead of a saddle tank" — the computer testing the wrong shape of tank. Hyundai and Kia have published a steady stream of ECM logic revisions for this code: Hyundai on the 2020 Sonata "to revise the logic for DTC P0456," Kia on the Sedona, Carnival and Sportage as service actions SA397, SA498 and others.

The cheap cause that gets skipped. The plug. Three separate Ford messages trace multiple EVAP codes including P0456 on 2020-2022 Explorer and Aviator models to "the wiring harness connection at connector C1198," and instruct: "Prior to replacement of any components, inspect the connector and wiring harness for any damage or pin fit concerns." Nissan traced P0456 on 2013-2014 NV200 vans to misrouted circuits at two power-distribution-module connectors.

Fix it yourself — prove the system holds before you accept "it's software." Ask for, or read yourself, the Mode 6 result for the 0.020-inch test against its limit, and ask whether the shop's leak test held pressure. "Software" is a satisfying answer and a cheap one, but it is only honest after a test that found nothing. The honest limit: you cannot install a manufacturer calibration yourself, and a generic reader cannot tell a marginal seal from a marginal calibration.

Where this does not apply. If there is a smell, a stall or a located leak, no calibration will fix it.

At the shop — say this: "Is there a TSB or calibration for P0456 on this exact VIN and engine? And before you reprogram, what did the leak test actually show?"

Route 5 · Your VIN is on a recall — the Ford and Volkswagen campaigns

The tell. A 2012-2018 Focus, or a 2015-2020 Golf, GTI, Golf Sportwagen, Jetta GLI or Audi A3 — especially one that stalls or will not start after refuelling.

What it actually is. Recall 18V-735, received by NHTSA on 22 October 2018, covers 1,282,596 2012-2018 Ford Focus vehicles with the 2.0L GDI or GTDI engine. Ford's stated defect is that "The Canister Purge Valve (CPV) may malfunction, possibly causing excessive vacuum in fuel vapor management system," and the consequence is written in the language of a safety recall, not an emissions one: "an engine stall while driving, without warning and without the ability to restart the vehicle, increasing the risk of a crash." The remedy reprograms the PCM, checks for fault codes, replaces the CPV as needed, and then "inspect and replace the carbon canister, fuel tank and fuel delivery module as necessary." Recall 19V-515 followed in July 2019 for 57,488 vehicles that "did not have the Powertrain Control Module (PCM) updated as intended."

On the Volkswagen side, recall 24V-110 covers 261,257 2015-2020 Audi A3 and Volkswagen Golf, GTI, Golf Sportwagen and Jetta GLI cars: "Due to a problem with a suction jet pump seal inside the fuel tank, fuel may flow into the evaporative emissions (EVAP) system and possibly leak out of the charcoal canister." It exists because NHTSA opened recall query RQ23-007 in December 2023 to test whether an earlier 2016 recall had actually worked.

The cheap cause that gets skipped. Running the VIN at all. None of these three campaigns has the string P0456 anywhere in its defect, consequence or remedy text, so searching the code will never surface them. Owners in our set found them by accident, by forum post, or not at all.

Where this does not apply. Being the right model and year is not eligibility, and this is the loudest complaint in the whole P0456 file: owner after owner reports the same symptoms, the same codes and the same recommended repair, and a VIN the campaign does not list.

At the shop — say this: "Run this VIN for every recall, emission recall, customer satisfaction program and extension, open or completed — and if a campaign was performed before, tell me the date and what was replaced."

Which P0456 repairs does somebody else pay for? Fewer than on most codes, and that is worth knowing before you get your hopes up. The Check Engine Code Index flags zero of the 269 manufacturer bulletins naming P0456 as a warranty extension or emissions campaign, which makes it the only code among its seventeen with no coverage document anywhere in the bulletin set. What exists instead is three safety recalls, none of which mentions the code in its text, plus the federal emissions warranty on emission control components, and a scattering of manufacturer service actions on Hyundai and Kia models. So the order that saves the most money is fixed: run the VIN for recalls and campaigns, check the in-service date against the emissions warranty, ask whether a bulletin exists for that exact engine, and only then start talking about parts. Pulscar's free check gives you a first read on the idle while you wait for the service desk to call back.

How often was it the cap? What 133 federal complaints say

This is the question the internet answers with a number nobody sources, so we went and counted.

Our sample: every vehicle complaint added to NHTSA's public files between 1 January 2015 and 8 September 2026, de-duplicated by report number — 777,040 unique owner reports. Matching the code whole-word and case-insensitive gives 133 complaints. Three more matched a loose text search and were discarded on reading: "PP0456", "P045600" and "P04567A". We read all 133 end to end.

Now read the bias, because it decides what the number means. People file federal complaints about problems that were expensive, unresolved or frightening. A P0456 that stopped coming back after somebody tightened a cap on a Tuesday essentially never produces one. So this is not evidence that the cap is rarely the cause. It is evidence that the cap is rarely the cause of a P0456 that is still unsolved weeks later — which is a narrower claim, and it happens to be the exact situation you are in if you are reading this page.

The practical reading is a sequence, not a probability. Check the cap because it is free and it takes two minutes. Then stop expecting it to be the answer, and ask for a test that locates the leak. The owners in this file who spent the most money are, almost without exception, the ones who bought a second and third cap instead of a smoke test.

What should you actually do first with a P0456, in order? Refit the cap properly and wipe its seal, because that costs nothing and four Mazda bulletins and two Nissan ones put it first. Run the VIN for recalls, because on a 2012-2018 Focus or a 2015-2020 Golf the answer is a free repair that does not mention the code anywhere. Write down every code stored beside it, because P1450 changes the diagnosis and P0451 says the sensor may be lying. Then drive normally for a week, because the monitor that has to clear this code is required to run only half as often as the large-leak one. Only after all four does it make sense to pay for a smoke test, and only after the smoke test does it make sense to buy a part. Pulscar's free check can tell you whether the idle itself is clean while you work through that list.

What 133 complaints naming P0456 turned out to be

Each of the 133 was given a single primary outcome — the thing the story turned out to be about, as the owner tells it.

The valve is the story. At 38 of 133 it is nearly as large as the next three categories put together, and it is the one category where the manufacturer documents and the owner complaints agree without qualification.

Almost nobody was stranded by the leak. Across all 133 reports the crash flag is set once, the fire flag never, and the towed flag once, against baselines of 4.37%, 1.82% and 5.11% across the full 777,040-complaint file. But 40 of the 133 mention a stall or a no-start, and nearly all of those are canister-purge-valve cars. The danger in this code is not the leak. It is the valve that also upsets the running engine.

Mileage. Sixty-five of the 133 state a usable mileage: the median is 79,000 miles, with the middle half between 57,226 and 109,500. This is not an old-car code.

What manufacturers themselves write: 269 bulletins, 44 problems

Complaints are what owners say. Bulletins are what the people who built the car tell their own dealers, and on this code there are a great many of them: 269 distinct manufacturer communications name P0456, from 11 manufacturer groups, the third-largest count of the seventeen codes in our Check Engine Code Index, behind only the two misfire codes.

192 of those 269 are Stellantis. We read the other 77 and grouped repeats — a Nissan and an Infiniti version of one bulletin, a revision, the same text reissued for more model years — into 44 distinct problems.

Finding 1: five of the 44 are instructions to test rather than parts to fit. Ford's SSM 50306 and SSM 46525, Kia's PS469 manual EVAP test "for hard to find/small leaks", Mazda's procedure for locating the leak, and an Audi bulletin whose entire purpose is to explain how to run an EVAP pressure test to diagnose P0456. On no other code we have analysed does the "how to find it" category come close to the parts categories.

Finding 2: six of the 44 name no repair at all. Two Audi bulletins say the light is on with no leak in the system. A Ford Bronco bulletin states the condition and stops. Two Hyundai documents list the codes and note "There are no drivability symptoms associated with these DTC(s)." One Volkswagen entry exists only to move the code to a different bulletin number. These are not evidence that nothing is wrong; they are evidence that the summary text does not say, and we show them as their own category rather than invent a cause.

Finding 3: the NVLD family is a geography lesson. Ten bulletins across Hyundai and Kia — grouped here as two problems, because each repeats one finding across model lines — describe salt particulate forming inside the natural vacuum leak detection sensor. Hyundai's wording is specific: the conditions "primarily occur in the California coastal areas that are subject to marine layer cloud cover." The remedy is an air filter kit, and an NVLD sensor if the kit is not enough. California does supply the largest share of our complaints, 21 of 133 — but it is also the largest car-owning state in the file and only two of the 133 are Hyundais, so we are not claiming the two facts are connected.

Finding 4: the newest document on this code was withdrawn. In February 2026 Nissan and Infiniti published bulletins for 2021-2026 Pathfinder, Murano, Rogue, Kicks, Armada, QX60 and QX80 vehicles with a smoke-test leak at the fuel filler tube lower seal, with a remedy that needs no parts at all: insert the supplied funnel completely into the filler tube and remove it five times, then re-test. On 30 May 2026 both bulletins were replaced by a one-line notice that they are "no longer active." If a dealer quoted you that procedure earlier in 2026, ask what replaced it.

The Stellantis asterisk: why "72% software" is not a fact about your car

The index reports that 194 of the 269 bulletins naming P0456 are software updates — 72%. That figure is correct and it is also the most misleading number on this page if you stop there.

178 of those 194 are Stellantis. Outside Stellantis, only 16 of 77 bulletins are software — 21%.

The reason is structural rather than mechanical. Stellantis publishes large numbers of engine-computer flash bulletins that each list dozens of diagnostic codes at once, so a code appears in the software column simply by being on a long list. Our index's own method measured this: the median flagged Stellantis bulletin names 7 distinct codes, 164 of them name ten or more, and one names 44.

What that means at the service desk. On a Jeep, Chrysler, Dodge or Ram it is genuinely worth asking whether a flash exists for your VIN, because the manufacturer publishes a lot of them. On a Mazda, a Kia, a Nissan or a Honda, "it's probably software" is a claim with very little published support behind it — four out of five non-Stellantis documents on this code describe something physical.

By make: what the file says about the ones people search for

Ford, 65 of 133 complaints including one Mercury. Twenty-nine are the Focus, then the Edge with 10, the Escape with 9, the Fusion with 6, the Mustang with 4. One story dominates: the canister purge valve, the vacuum it pulls on the tank, the stall that follows, and recall 18V-735. The second is the capless filler, which Edge and Mustang owners describe as a problem with no $20 answer — there is no cap to replace, so the quote is the whole filler pipe.

Stellantis brands, 30 complaints. Jeep 12, Chrysler 10, Dodge 6, Ram 1, Fiat 1, with the Grand Cherokee leading at 8. Fewer stalls than the Ford group, more owners reporting that a shop replaced a valve, then another valve, and the code came back. Given how many flash bulletins Stellantis publishes here, the bulletin search is worth doing first on these cars.

Hyundai and Kia, 9 complaints, and the largest documented fix. The NVLD salt family and the ECM logic revisions are both live programmes, and the Kia bulletins are written as service actions with production date ranges. If you own a Sonata Hybrid, Ioniq, Optima, Optima Hybrid or Niro from those years, the words to use are air filter kit and NVLD.

Nissan and Infiniti, 3 complaints, and the best-written bulletin. The vent control valve O-ring document is the clearest example on this page of a manufacturer naming a $10 part instead of a $200 one, and it is explicit that the cap must be ruled out first.

Mazda, Toyota, Honda and Volkswagen. Mazda's position is the cap, four times over. Toyota's Corolla bulletin is a filler neck pretending to be a cap. Honda's is a capless flap pretending to be a fill pipe. Volkswagen's is recall 24V-110 plus software updates for the Jetta and Taos — and VW owners in our set are the ones most often told that "topping off" caused it, which four visits did not make true for one Atlas Cross Sport owner.

Clearing P0456 before an inspection: why it is worse than useless

Clearing the code the morning of the test fails for three documented reasons, and on this code the third one is unusually punishing.

The light is an automatic fail. Under 40 CFR 85.2222, the federal procedure for OBD inspections, a vehicle whose check engine light is commanded on "shall fail the OBD inspection."

A freshly cleared car is not ready. The same rule allows 2001 and newer vehicles no more than one unset readiness monitor. Clearing the code resets the monitors.

And this monitor is the slowest one you own. The 0.020-inch leak monitor has the lowest required in-use performance ratio in the regulation, 0.260 against 0.520 for the large-leak and purge-flow tests. Clear the code and you are waiting for the least frequent test on the car to run before the station will look at you — and if the leak is real, it fails again as soon as it does run.

The honest route to a pass is the repair, then a week of ordinary driving, then a readiness check with your own reader before you join the queue.

What actually happens at the shop, checkpoint by checkpoint

Checkpoint 1 — the free two. "Has the cap and filler seal been checked, and has the VIN been run for recalls and campaigns?" On a Focus or a Golf these two questions can end the conversation.

Checkpoint 2 — the full code list. "Which codes are stored, pending and in history?" P1450 moves you to the valve and the recall; P0451 or P0452 says the sensor may be lying.

Checkpoint 3 — the leak test, and where it pointed. "Did the smoke or pressure test find the leak, and where exactly?" A test that found nothing is a result too, and it should be written down before anybody reprograms anything.

Checkpoint 4 — the valve question, asked Ford's way. "Was the purge valve leak-tested, or swapped on suspicion?" Ford's own message says its returned valves tested fully functional.

Checkpoint 5 — the number and the ceiling. "What is the Mode 6 result for the 0.020-inch test against its limit, what is the diagnostic charge, and will you call me before anything over this figure?" The overcharging guide covers how quotes inflate.

Three diagnostic traps that cost owners the most

Trap 1: the second and third gas cap. The situation: the light came back a week after a new cap. What the shop says: "Must be a bad cap. Try another one." What is really happening: the cap is one of several possible leaks, and the threshold is cumulative. Three owners in our 133 were on their third cap; one had a clean smoke test in hand at the time. Price versus bill: $50 to $150 for a smoke test that locates the leak, against an unbounded number of $20 caps and the diagnosis you still have not had. Your protective question: "Before another cap, will you smoke-test the system and show me where it leaks?"

Trap 2: the purge valve fitted on suspicion. The situation: P0456 alone, no smell, no stall, and a $300 quote off the code. What the shop says: "It's always the purge valve on these." What is really happening: it often is — 38 of our 133 — but Ford tested returned valves from 21 models and found them fully functional, and asks for a leak test first. A valve fitted without a test is a coin flip you paid for. Price versus bill: a leak test, or a replaced valve and the same light a fortnight later. Your protective question: "Which test condemned the valve, and what did it show?"

Trap 3: "it's just software" with no test behind it. The situation: nothing found, and an offer to reprogram. What is really happening: software genuinely is the answer sometimes, and Audi and Ford both publish bulletins saying so. But 72% of bulletins naming this code being software is a fact about Stellantis flash documents, not about your Mazda: outside Stellantis the figure is 21%. Price versus bill: free inside warranty, but a reprogram that hides a real leak means you pay twice and fail the inspection anyway. Your protective question: "Is there a calibration for this exact VIN and engine, and what did the leak test show before you decided on software?"

Three quotes, decoded

These are composite scenarios, not individual customers — the mileages and outcomes are typical of the pattern each describes, assembled so the arithmetic is visible. We do not publish invented testimonials.

Scenario 1 — the Focus that was not a leak at all. A 2014 Focus at 78,000 miles, P0456 with P1450, stalling after fill-ups. The first independent quote is a purge valve and a canister. The owner runs the VIN, finds recall 18V-735, and the dealer reprograms the PCM, replaces the canister purge valve and checks the tank and fuel delivery module at no charge, because the recall is about the stall. Lesson: on this code the recall lookup is a diagnostic step, not paperwork.

Scenario 2 — the Nissan that needed a ten-dollar part. A Sentra at 67,000 miles with a lone P0456. The cap checks out on a leak test. The technician follows NTB17-082, finds the leak where the vent control valve meets the canister, and fits the O-ring rather than the valve, exactly as the bulletin instructs. Lesson: ask which part of the valve failed, not just which valve.

Scenario 3 — the third cap that should have been a smoke test. A 2015 Sienna at 96,000 miles. Two new caps, two returns of the light, and a smoke test that finds nothing at the filler. The third visit tests the whole system instead of the cap, and the leak is a seepy joint further back. Lesson: when a cheap part fails twice, the next thing to buy is a test, not a third part.

Should you buy a used car with a P0456 history?

Ask what was actually done, and get it in writing. A P0456 answered by a located leak and a named repair is a finished job. A P0456 answered by a cap, then a valve, then another valve is an unfinished one, and the next owner inherits it.

On a 2012-2018 Focus, check the recall status before anything else. 18V-735 covers 1,282,596 cars and the consequence is a stall without warning. If the recall shows incomplete, that is a free dealer visit and a reason to negotiate, not a reason to walk. If it shows complete and the car still stalls after refuelling, 19V-515 exists precisely because some cars did not get the update they were supposed to.

On a 2015-2020 Golf, GTI, Golf Sportwagen, Jetta GLI or Audi A3, check 24V-110 the same way — and remember that recall exists because NHTSA queried whether the 2016 version had worked.

What to pay for at the inspection. Stored and pending codes printed, the Mode 6 result for the 0.020-inch test, and a note of which readiness monitors are complete. A seller who cleared the codes yesterday cannot hide that last one: the slowest monitor on the car will still be incomplete.

The one combination to walk away from is a fuel smell with a full tank plus a history of replaced EVAP parts. That is Route 3 with somebody else's money already spent on Routes 1 and 2.

After the fix: verify, then prevent

Verify the monitor, not the light. Drive a week of ordinary trips, then check with your own reader that the evaporative monitor shows complete and that P0456 has not returned as pending. A dark dashboard on day two means nothing here.

Keep the paperwork, and keep the test. The invoice should say where the leak was found, not just what was replaced. On a car with a recall history that document is what a future argument turns on.

Prevent the cheap version. Turn the cap until it clicks, keep the seal and the filler lip clean, do not force the pump past the first click, and look at the filler area whenever the car is on a lift. Chafed lines and cracked canister mountings are found by looking.

Record a baseline. Thirty seconds of the engine at idle, saved now, turns the next rough idle into a comparison instead of an argument. The recording guide has the method, and the check engine light guide covers the next light.

Sources and documents

Primary documents only. Where a claim is marked 🟢 it rests on an item below, opened and read, with every population, date and quoted phrase taken from the document itself. Claims marked ⚪ are estimates or owner reports.

1. 13 CCR 1968.2, California OBD II — section (e)(4.2.2)(C) (0.020 inch, "cumulatively"), (e)(4.2.1) (the O'Keefe Type B orifice, part B-20-SS), (d)(3.2.1)(B) (in-use ratios 0.260 and 0.520), (g)(4.5.4)(F) (separate 0.020-inch test results), (e)(4.2.3) (tanks over 25.0 gallons). Read 19 September 2026. Regulation
2. 40 CFR 85.2222 — OBD inspection and readiness rules. Regulation
3. NHTSA recall 18V-735 / Ford 18S32, received 22 October 2018 — 1,282,596 2012-2018 Focus 2.0L GDI or GTDI; canister purge valve; "engine stall while driving, without warning and without the ability to restart the vehicle." Recall record
4. NHTSA recall 19V-515 / Ford 19S22, 5 July 2019 — 57,488 vehicles previously recalled under 18V-735 whose PCM "did not have the Powertrain Control Module (PCM) updated as intended." Recall record
5. NHTSA recall 24V-110 / VW 20UF, Audi 20YF, 14 February 2024 — 261,257 2015-2020 Audi A3 and VW Golf, GTI, Golf Sportwagen, Jetta GLI; suction jet pump seal inside the fuel tank. Recall record
6. NHTSA recall query RQ23-007, opened 12 December 2023, closed 1 July 2024 — remedy effectiveness and scope of VW recall 16V-647, which produced 24V-110. ODI investigation file layout
7. Ford SSM 50306, 23 November 2021 — 21 Ford and Lincoln models; "Ford has tested returned EVAP canister purge valves and found them to be fully functional. Prior to replacing the canister purge valve or other EVAP components, perform the following tests."
8. Ford SSM 46525, 13 April 2017 — 2017 Escape and MKC; check for liquid fuel in the vapor lines, which "should not be present," before replacing the purge valve.
9. Ford TSB 17-0016, 22 February 2017 — 2016-2017 Focus RS; "an intermittent sticking evaporative emission canister purge valve."
10. Ford TSB 19-2207, 28 June 2019 — 2013-2017 C-Max, Fusion and MKZ hybrids; "the sealing capability of the canister purge valve"; PCM software plus the valve.
11. Ford TSB 21-2020, 27 January 2021 — 2020 Escape 2.0L FWD built on or before 20 March 2020; calibration "configured for an L-tank instead of a saddle tank."
12. Ford TSB 21-2106, 8 April 2021 — 2017-2021 Transit; "a kinked evaporator emissions fresh air vent line"; remedy is a zip tie.
13. Ford TSB 25-2308, 20 August 2025 — 2022-2024 Maverick; "a chaffing or rubbing condition between the fuel vapor canister line and the air vent hose."
14. Ford TSB 24-2086 and 25-2112 — 2021-2023 Bronco 2.7L and 2022-2024 Explorer; "the software level of the PCM and has no effect on vehicle operation."
15. Ford SSM 50348, 51132 and 51247, 2021-2022 — 2020-2022 Explorer and Aviator; wiring connector C1198, inspected before any part is replaced.
16. Ford SSM 49313, 6 November 2020 — 2020-2021 Escape and Corsair; carbon canister damage "which may have occurred during vehicle transportation."
17. Nissan NTB17-082 and Infiniti ITB17-042, latest 7 February 2022 — "DTC P0456 CAUSED BY EVAP LEAK AT VENT CONTROL VALVE O-RING ... DO NOT REPLACE the vent control valve."
18. Nissan NTB13-097 and Infiniti ITB13-044 series, latest 7 February 2022 — all model year 2012 and newer; "a loose fuel fill cap may be the cause."
19. Nissan NTB26-006 and Infiniti ITB26-002, 10 February 2026, withdrawn 30 May 2026 — fuel filler tube lower seal; funnel procedure.
20. Mazda 01-005-19-3548, 01-008-15-3045, 01-013-18-3478 and 01-014-14-2932, 2014-2019 — "most likely the result of the customer not getting the fuel cap completely closed after fueling the vehicle."
21. Mazda SA-030-18, 25 June 2018 — 2017 Mazda3; "leakage detected from the catch tank unit (1) at the purge valve (2)."
22. Hyundai 19-01-030H, 20-01-020H, 20-FL-002H, 20-FL-003H, 21-01-035H, 21-FL-002H and Campaign T4C — salt inside the natural vacuum leak detection sensor; air filter kit and NVLD sensor.
23. Hyundai 5NPY6N5B02, 5 January 2017 — 2017 Santa Fe Sport; "the fuel filler cap seal may not provide proper sealing."
24. Hyundai 5NPU7G9E02 and 5NP-Z2J3V-04, 2017 — 2011-2016 Veloster and Sonata Hybrid; "a potential crack found at the evaporative canister insert nuts."
25. Kia TSBFUE040, 18 April 2018 — 2012-2017 Rio; canister close valve "stuck open."
26. Kia PS322, 22 May 2014 — 2011-2015 Optima; "closely inspect the mounting holes of the charcoal canister for visible cracks."
27. Kia PS469, 18 January 2017 — ten Kia models; manual EVAP test for hard-to-find or small leaks after a not-found self-test.
28. Kia PS567 series, latest 20 November 2020 — 2017 Cadenza and Optima Hybrid; "DTC P0456/P0442 WITH FUEL PUMP SEAL DAMAGE."
29. Honda ATS170206 and Acura BTS170202, 18 October 2017 — 2016-2017 Pilot and Ridgeline, 2017 MDX and NSX; "don't replace the fuel fill pipe! Instead, replace the mis-fuel inhibit device and the main flap unit."
30. Toyota T-SB-0086-16, 28 July 2016 — 2014-2016 Corolla; "No Click" cap condition traced to a filler neck out of position.
31. Audi bulletins A011510 and A011651, 2015-2016 — Q7; "The MIL is on due to DTC P0456 being stored in the ECM. There is no leak in the EVAP system."
32. Subaru 11-152-14, 15 December 2014 — 2015 Legacy and Outback; clamps added to the canister purge control solenoid valve hose.
33. Our P0456 complaint analysis — all 133 complaints naming P0456 read in full and classified; the cap counted in every direction; mileage, companion codes and structured outcome flags. Method · all 133 classified (CSV) · by make and model (CSV) · model year and report year (CSV) · CC BY 4.0
34. Our P0456 bulletin analysis — the 77 non-Stellantis bulletins from the published index list, read and grouped into 44 distinct problems with a quotation each. Distinct problems (CSV)
35. Pulscar Check Engine Code Index — 269 bulletins naming P0456 across 11 manufacturer groups, at least 194 software updates, zero coverage campaigns, and the complaint and mileage figures this page builds on. Index · code index (CSV) · bulletin list (CSV)
36. Our complaint sample — 777,040 unique complaints added 1 January 2015 to 8 September 2026, de-duplicated by report number. NHTSA complaint field layout · TSB field layout

What we looked for and did not find. No Part 573 safety recall has the string P0456 in its defect, consequence, remedy or notes text — the three recalls above are reached through the component and the narrative, which is why owners kept finding them by accident. We found no warranty extension or emissions campaign among the 269 bulletins, the only code in our index with none. We did not test whether California's coastal climate actually produces more P0456 reports, because we have no per-state vehicle denominator and will not imply one. And we could not re-price this repair in this pass: the estimator sites that publish these ranges refuse automated access, and we do not work around that.

EVAP and fuel: P0455 large leak, P0442 small leak, gas smell, fuel pump cost. Codes and lights: check engine light, code index, P0420, P0430, P0171, P0300, most expensive codes. Symptoms: stalls while driving, starts then dies, rough running. Money: diagnostic costs, warranty extensions, our story.

Methodology. Regulation text, recall populations, bulletin wording and every quoted phrase come from the documents above, read on 19 September 2026. Complaint and bulletin counts come from NHTSA's public files, de-duplicated; they are owner reports and manufacturer communications, not audited repairs and not failure rates. Price ranges are estimates carried forward from our cost guides.

Prices verified: September 2026 · Technical documents verified: September 2026 · Reviewed quarterly

Had a P0456 that took more than one visit? Email [email protected] with the year, engine, mileage and what finally located the leak — owner data is how these ranges get sharper.