PulsCar analysis of NHTSA public data, September 2026 Rough idle: what 5,064 owner complaints and 966 manufacturer bulletins say Licensed CC BY 4.0. Query run 18 September 2026; complaint set and every figure derived from it recomputed 18 September 2026 (see CORRECTION below). -------------------------------------------------------------------------- CORRECTION, 18 September 2026 -------------------------------------------------------------------------- The complaint set published earlier on this page (2,338) was built with the word lists below matched against the narrative WITHOUT case folding -- the line "roughness words (regex, narratives are uppercase)" says so, and that parenthetical is only true through 2020. NHTSA narratives are mixed case from 2021 on, so the query kept almost nothing received after 2020: a rebuild of the same rules matched 144 complaints in 2021 and 5, 6, 4, 3 and 5 in 2022-2026, against roughly 300-600 a year before that. Re-running the rules case-insensitively returns 5,064 complaints and restores the recent years. Every complaint figure on this page, in the article and in rough-idle-by-make.csv, rough-idle-top-models.csv and rough-idle-what-owners-said-was-repaired.csv has been recomputed from that set. NOT affected, and left alone: the 966 manufacturer bulletins and the seven recall campaigns. Those queries run against NHTSA's bulletin and recall text, which is mixed case across the whole window; re-running them case-insensitively returns the same answers (a case-sensitive bulletin match returns 157, which is not what was published). Two caveats about this correction, both real: 1. The rebuild re-implements the rules documented below from this file. Run case-sensitively it returns 2,756 complaints against the 2,338 published earlier -- 18% broader -- so it is a faithful reconstruction of the documented rules but not of the original script, whose sudden- acceleration and launch-shudder exclusions evidently caught more than the descriptions below reproduce. The rebuilt set's towed, crash and fire shares are correspondingly higher than the ones published earlier (2.65%/3.63%/1.02% case-sensitively, against 1.03%/1.28%/0.47%), and the mis-sort rate measured in the SAMPLE READING below belongs to the ORIGINAL query, not to this one. Treat the flag shares as the weakest numbers on the page. 2. The phrase-subset and repair-category regexes were re-derived from the descriptions below, so those counts are new measurements of the same intent rather than the identical measurement. The air-conditioning subset now uses (AIR CONDITION*|A/C|AC). -------------------------------------------------------------------------- SOURCE FILES -------------------------------------------------------------------------- NHTSA ODI complaint flat files COMPLAINTS_RECEIVED_2015-2019, _2020-2024, _2025-2026 https://static.nhtsa.gov/odi/ffdd/cmpl/ (field dictionary: CMPL.txt) 777,040 distinct vehicle complaints (ODINO, product type V) received 1 January 2015 to 8 September 2026. NHTSA manufacturer communication (TSB) flat files TSBS_RECEIVED_2015-2019, _2020-2024, _2025-2026 https://static.nhtsa.gov/odi/ffdd/tsbs/ (field dictionary: TSBS.txt) Fields used: 4 TSB/Document ID, 5 Mfr Communication Date, 8 Make, 9 Model, 10 Model Year, 14 Summary. NHTSA recall flat file FLAT_RCL.zip (campaign, make, model, year, component, potentially affected, defect, consequence, remedy). -------------------------------------------------------------------------- WHICH COMPLAINTS COUNT (the rough-idle set: 5,064) -------------------------------------------------------------------------- A complaint counts when its narrative contains ONE SENTENCE in which a roughness word and an idle/stationary word appear within 60 characters of each other. roughness words (regex, CASE-INSENSITIVE -- printed in capitals for readability only; see CORRECTION): ROUGH | ROUGHLY | SHAK* | VIBRAT* | SURG* | HUNT* | FLUCTUAT* | MISFIR* | SPUTTER* | STUMBL* | JUDDER* | SHUDDER* | CHUG* | BOUNC* idle / stationary words: IDLE | IDLES | IDLING | IDLED | STOP LIGHT | STOPLIGHT | RED LIGHT | STOPPED AT | AT A STOP | AT A COMPLETE STOP | IN PARK | IN NEUTRAL | DRIVE-THRU (and spelling variants) | WHILE STOPPED | COMES TO A STOP | AT A LIGHT | AT THE LIGHT | SITTING STILL | STANDING STILL EXCLUSIONS, applied per sentence: denials and negatives -- DOES NOT IDLE ROUGH, NO ROUGH IDLE, NOT ROUGH, DOESN'T IDLE ROUGH, NEVER IDLED ROUGH, SMOOTH IDLE, IDLES SMOOTH; road-surface uses of "rough" -- ROUGH ROAD*, ROUGH TERRAIN, ROUGH PAVEMENT, ROUGH SURFACE, ROUGH PATCH, ROUGH SPOT, BUMPY, POTHOLE*, ROUGH GRAVEL, ROUGHLY , and the phrase IDLE AIR VOLUME (a Nissan relearn procedure name, not a symptom); sudden-acceleration narratives -- SURGED FORWARD / AHEAD / BACKWARD, SURGED FEET, LURCHED FORWARD, SUDDEN(LY) ACCELERAT*, UNINTENDED ACCELERAT*, ACCELERATED ON ITS OWN, TOOK OFF ON ITS OWN; launch-shudder narratives (the car shudders pulling away, which is a different symptom on a different page) -- TURNS/TURNED GREEN, PULLING AWAY, FROM A (DEAD) STOP, TAKING OFF, PRESSED/STEPPED ON THE GAS or ACCELERATOR, GAVE (IT) GAS, APPLIED THE GAS, GAS PEDAL APPLIED, GO INTO DRIVE; "IN DRIVE" was deliberately dropped from the idle word list after sample reading: it matched stall-while-driving narratives ("cut off while still in drive"). "IN PARK" and "IN NEUTRAL" were kept because they imply a stationary vehicle. Makes excluded: motorcycle, heavy-truck, bus and RV manufacturers (Harley-Davidson, Indian, Triumph, Ducati, Kawasaki, Yamaha, Suzuki, Polaris, Can-Am, KTM, Vespa, Piaggio, Aprilia, Moto Guzzi, Arctic Cat, Royal Enfield, Husqvarna, Zero, BMW Motorrad, Mack, Hino, Kenworth, Peterbilt, Paccar, International, IC Bus, Freightliner, Western Star, Blue Bird, Thomas Built, Navistar, Prevost, Workhorse, Spartan, Autocar, Forest River, Volvo Trucks, Volvo Bus, Isuzu Trucks, UD Trucks, Sterling, Gillig, Chalet, Bison). SAMPLE READING (required before publishing any keyword count) Three rounds of 30 randomly drawn matches were read in full. Round 1 (first draft of the query, 2,802 matches): 3 of 30 were not rough-idle reports -- 10%. Round 2 (after dropping "IN DRIVE", 2,722 matches): 8 of 30 were not rough-idle reports -- 27%. The misses were sudden-acceleration reports and launch shudders, so the two exclusion lists above were added. Round 3 (final query, 2,338 matches): 4 of 30 were not rough-idle reports -- 13%. The survivors were a stall-while-driving narrative, a unintended-acceleration report phrased as "surged 12 feet", a "rough patch of road" at a stop light, and one report where the roughness was not at idle. The first three were then denied by rule (ROUGH PATCH, SURGED FEET), which removed 2 more complaints from the set. TREAT EVERY COUNT ON THIS PAGE AS CARRYING ROUGHLY ONE MIS-SORT IN EIGHT. -------------------------------------------------------------------------- MILEAGE AND RATE FIELDS -------------------------------------------------------------------------- Mileage: NHTSA's MILES field, rows with 0 < miles < 400,000 only. 3,017 of the 5,064 complaints report a usable mileage. Median 70,000; first quartile 32,000; third quartile 107,000. 27.5% at or under 36,000 miles; 42.0% at or under 60,000; 30.2% at or above 100,000. All-complaint baseline for the same file and exclusions: median 70,000 over 392,935 complaints reporting a usable mileage. Towed / crash / fire: NHTSA's TOWED, CRASH and FIRE flags, share answering Y. Rough-idle set: towed 3.89%, crash 3.24%, fire 0.85%, 137 injuries, 0 deaths. (See caveat 1 in CORRECTION before quoting these three.) All-complaint baseline (777,040): towed 5.11%, crash 4.37%, fire 1.82%. Phrase subsets are share of the 5,064 whose FULL narrative matches: stall* 1,080 (21.3%); check engine light / MIL 1,491 (29.4%); misfir* 653 (12.9%); spark plug or coil 396 (7.8%); neutral 275 (5.4%); throttle body / idle air control / IAC 182 (3.6%); air conditioning 162 (3.2%); injector 169 (3.3%); cold start 97 (1.9%); engine or motor mount 54 (1.1%); vacuum leak / intake gasket / PCV 51 (1.0%); EGR 19 (0.4%). All-complaint baselines computed the same way (recomputed case-insensitively 18 September 2026; the earlier 3.48% / 2.20% / 1.07% were uppercase-only matches over the same 777,040 rows): stall* 7.63%, check engine light / MIL 6.58%, "no fault found" family 1.86%. "No fault found" family (201 complaints, 4.0%): could not duplicate / replicate / reproduce / find, unable to duplicate, no problem/fault/issue/ trouble found, operating as designed, within specification, normal characteristic or operation, nothing found, no codes. -------------------------------------------------------------------------- WHAT OWNERS SAID WAS REPAIRED (1,457 complaints) -------------------------------------------------------------------------- A complaint enters this table when its narrative contains a repair verb (DIAGNOS*, DETERMINED, REPLACE*, REPAIR*, FIXED, INSTALL*, NEEDED A NEW, TOLD ME/US THE) AND names at least one part family. Categories overlap, so the percentages do not sum to 100. This is NOT a cause ranking: it counts which rough idles ended in a federal safety complaint, and people file with the government over a four-figure engine bill, not over a $90 plug set. Category regexes are in rough-idle-what-owners-said-was-repaired.csv order: engine internals PISTON RING | VALVE SEAT/GUIDE/SPRING/STEM | BURNT VALVE | CYLINDER HEAD | HEAD GASKET | CAMSHAFT | TIMING CHAIN/ BELT | LOW COMPRESSION | SHORT/LONG BLOCK | NEW ENGINE | ENGINE REPLACED/REPLACEMENT | LIFTER | ROCKER ARM | VTC ACTUATOR ignition SPARK PLUG | IGNITION COIL | COIL PACK | IGNITION WIRE | PLUG WIRE | DISTRIBUTOR air metering THROTTLE BODY | THROTTLE PLATE | IDLE AIR CONTROL | IAC | MASS AIR FLOW | MAF | THROTTLE POSITION SENSOR fuel FUEL INJECTOR | INJECTOR | FUEL PUMP | FUEL FILTER | FUEL RAIL | HIGH PRESSURE FUEL electrical BATTERY | ALTERNATOR | GROUND WIRE/STRAP | WIRING HARNESS software SOFTWARE UPDATE/UPGRADE | REPROGRAM* | REFLASH | ECM/PCM UPDATE | CALIBRATION UPDATE | UPDATED THE SOFTWARE vacuum VACUUM LEAK/HOSE/LINE | INTAKE (MANIFOLD) GASKET/LEAK/ BOOT | PCV | PURGE VALVE | EVAP driveline TORQUE CONVERTER | TRANSMISSION REBUILT/REPLACED | VALVE BODY | CLUTCH PACK | CVT sensors OXYGEN/O2 SENSOR | CRANK POSITION SENSOR | CAM POSITION SENSOR | COOLANT TEMPERATURE SENSOR | KNOCK SENSOR mounts ENGINE MOUNT | MOTOR MOUNT | TRANSMISSION MOUNT carbon CARBON BUILD-UP/DEPOSIT/CLEAN | WALNUT BLAST | INTAKE VALVE CLEAN EGR EGR -------------------------------------------------------------------------- MANUFACTURER BULLETINS (966) -------------------------------------------------------------------------- A manufacturer communication counts when its Summary field matches: ROUGH IDL* | IDLE (IS/RUNS/FEELS) ROUGH | IDLES ROUGH | IDLE VIBRATION | VIBRATION AT IDLE | VIBRATION WHILE/WHEN IDLING | IDLE SURGE/SURGING/ FLUCTUAT*/HUNT* | UNSTABLE IDLE | ERRATIC IDLE | ENGINE SHAKES AT IDLE | SHAKE AT IDLE | SHUDDER AT IDLE | POOR IDLE | IDLE QUALITY | ROUGH RUNNING | RUNS ROUGH | MISFIRE AT IDLE | STALLS AT IDLE One bulletin = one NHTSA document ID (the flat file carries one row per make/model/year, so counting rows would multiply a single bulletin by its vehicle list -- 34,184 rows collapse to 966 documents). The same heavy-truck, bus and motorcycle makers are excluded, plus Cummins. 966 distinct bulletins. 327 (33.9%) mention software, reprogramming, reflash, calibration or a PCM/ECM update. 29 mention a warranty extension, warranty enhancement, customer satisfaction program, special coverage, service campaign or emissions recall. Per-manufacturer-group counts and software shares are in rough-idle-bulletins-by-manufacturer.csv. Groups follow the same convention as the PulsCar Check Engine Code Index: GM (Buick, Cadillac, Chevrolet, GMC, Pontiac, Saturn, Hummer, Saab, Oldsmobile, Geo); Stellantis (Chrysler, Dodge, Jeep, Ram, Fiat, Alfa Romeo, SRT, Maserati); Ford (Ford, Lincoln, Mercury); Honda (Honda, Acura); Toyota (Toyota, Lexus, Scion); Nissan (Nissan, Infiniti); Hyundai (Hyundai, Genesis); VW Group (Volkswagen, Audi, Bentley, Lamborghini); JLR (Jaguar, Land Rover); every other make stands alone. A bulletin that covers several groups is counted once in each group's row, so the group column sums to more than 966. -------------------------------------------------------------------------- RECALLS -------------------------------------------------------------------------- Recall campaigns whose defect or consequence text names a misfire, rough running, a rough idle or a stall at idle: 7 distinct campaigns in the whole 2015-2026 file, of which two are passenger vehicles (18V-636, Alfa Romeo, 34,357 vehicles; 26V-493, Genesis G80, 42 vehicles) and the rest are motorcycles, scooters, buses and medium-duty trucks. This is a finding, not a gap in the query: a rough idle is a driveability complaint, and a recall requires an unreasonable risk to safety. -------------------------------------------------------------------------- LIMITATIONS -- READ BEFORE QUOTING ANY NUMBER -------------------------------------------------------------------------- 1. Complaints are self-reported, unverified, and voluntary. They are not failure rates and cannot be turned into one, because NHTSA publishes no denominator: nobody knows how many of each model are on the road. 2. Counts therefore favour vehicles that sell in large numbers and vehicles whose problem has been written about. Nothing here says any vehicle is more or less reliable than another, and no honest reading of complaint counts can say that. 3. The keyword method carries roughly one mis-sort in eight (see SAMPLE READING). Small counts -- anything under about 30 -- should be read as "a handful of owners", not as a rate. 4. Mileage, towed, crash and fire fields are the owner's own entry. The towed and crash shares of any complaint set are inflated relative to the real world, because a frightening event is what prompts a filing. 5. Bulletin text in the flat file is NHTSA's summary of the manufacturer's document, not always the manufacturer's own wording. Where this analysis quotes a bulletin word for word, the quoted words are from the summary record or, where a scanned PDF exists on static.nhtsa.gov, from the PDF; the article says which. 6. "Mentions software" counts the word, not the outcome. A bulletin that mentions reprogramming may also replace hardware. -------------------------------------------------------------------------- FILES -------------------------------------------------------------------------- rough-idle-by-make.csv make, complaints, miles_reported, median_miles, pct_at_or_under_36000_miles, pct_towed, pct_mentioning_stall. Makes with fewer than 5 complaints omitted. rough-idle-top-models.csv make, model, complaints, top_model_year, complaints_in_top_year, and the same mileage/towed/stall columns. Top 60 models, 5-complaint floor. rough-idle-what-owners-said-was-repaired.csv repair_category, complaints_naming_it, pct_of_1457_that_name_a_repair. rough-idle-bulletins-by-manufacturer.csv manufacturer_group, idle_bulletins, mention_software_or_reprogramming, pct_software, mention_warranty_extension_or_campaign. Groups with fewer than 3 bulletins omitted. No personal data is published: no VIN, no city, no dealer name, no owner free text beyond short quoted snippets in the article itself.