Nearly One in Four Used-Car Recalls Is a Failed Federal Standard, Not a Broken Part

Key facts

• 5,294 car and light-truck recall campaigns in NHTSA's post-2010 Part 573 file, covering 489,645,634 vehicles as filed.
• 1,269 (23.97%) cite a specific Federal Motor Vehicle Safety Standard. They cover 58,614,068 vehicles - 11.97% of the total, so noncompliance recalls are numerous but small.
• 43 distinct standards appear. The top three by campaign count: FMVSS 108 lamps (201), FMVSS 208 occupant crash protection (189), FMVSS 111 rear visibility (156).
• By vehicles, FMVSS 111 rear visibility leads at 16,192,156 - about 2.4x FMVSS 108's 6,754,870 - and it is not one big filing: the largest single campaign is 7.6% of it.
• 84 of the 156 rear-visibility campaigns name software or a display function, carrying 9,016,635 vehicles (55.69%).
• A noncompliance campaign covers a median of 3,184 vehicles against 5,128 for a defect campaign.
• Restricting to cars and light trucks matters: the unrestricted file gives 2,802 campaigns and 69,269,667 vehicles, so the campaign count more than halves while the vehicle count falls only 15.4%.

What is the difference between a defect recall and a noncompliance recall?

A defect recall is filed because something in the vehicle is failing or will fail. A noncompliance recall is filed because the vehicle, as built, does not meet a Federal Motor Vehicle Safety Standard - the binding construction rules in 49 CFR Part 571. The manufacturer files the same Part 573 report either way, NHTSA assigns the same kind of campaign number, and the vehicle shows the same open recall on a VIN lookup.

The distinction is recorded, though. Field 19 of NHTSA's recall flat file carries the FMVSS standard number when the campaign is a noncompliance filing. Of 5,294 car and light-truck campaigns, 1,269 carry one.

How often is a recall a standards failure rather than a broken part?

Nearly one in four. On the car and light-truck basis, 1,269 of 5,294 campaigns - 23.97% - cite a standard.

That is the count. The vehicle totals point the other way: those 1,269 campaigns cover 58,614,068 of 489,645,634 vehicles, or 11.97%. Noncompliance recalls are about twice as common as their share of vehicles suggests, because they are smaller.

Which federal standard do carmakers break most often?

Lighting. FMVSS 108, "Lamps, reflective devices, and associated equipment", accounts for 201 of the 1,269 campaigns - more than any other standard, and more than one in six.

But the standard that has put the most cars under recall is a different one, and the gap is wide.

Standard titles are as published in 49 CFR Part 571. The full table of all 43 standards is in the frozen dataset.

Why does rear visibility cover so many more vehicles?

Because the rule is new, it applies to every light vehicle, and what it regulates is a picture rather than a part. FMVSS 111 has required a rear-visibility image on new light vehicles since May 2018, and a campaign is filed whenever that image fails to behave as the standard requires.

Two things in the data make this concrete. First, the exposure is broad rather than one bad filing: across 156 campaigns the largest is 19V654000, a Nissan and Infiniti campaign of 1,228,830 vehicles, which is only 7.6% of the standard's total, and the median campaign is 18,411 vehicles. Second, it is mostly software. Of the 156 campaigns, 84 name software or a display function as the component, and those carry 9,016,635 vehicles - 55.69%.

The largest rear-visibility campaigns, with the filer as recorded in the Part 573 report:

Which manufacturers file the most noncompliance recalls?

Ford and Chrysler, at 177 campaigns each, with Chrysler covering almost twice as many vehicles.

This table required normalising the filer name and the correction was large. NHTSA's MFGNAME field carries several spellings of the same company: General Motors appears as "GENERAL MOTORS CORP.", "GENERAL MOTORS LLC", "General Motors LLC" and "General Motors, LLC", and reading them as four filers understates it at 69 campaigns instead of 120. Chrysler appears under five spellings, including one carrying "(Stellantis)", and goes from 121 to 177. These counts are not a ranking of how compliant a manufacturer is - a company that sells more vehicles in more variants files more of everything.

What a VIN check can and cannot tell you here

It can tell you that a recall exists on the specific car you are looking at, and how many. That is the part most buyers skip, and it is the part a VIN check is for. Zilocar also shows the car's accident and damage records including airbag deployment, salvage auction records, odometer readings and rollback indications, theft records, ownership history and sales-listing history.

It cannot tell you whether a given recall is a defect filing or a noncompliance filing. That distinction is in the manufacturer's Part 573 report, and the place to read it is NHTSA. A VIN check also cannot tell you whether the remedy has been performed, whether a software update has actually been flashed to that car, or whether the vehicle is caught in an open NHTSA investigation that has not yet become a recall. For remedy status, use NHTSA's own VIN tool or ask a franchised dealer - they see the repair record, and a vehicle history report does not.

What to do with this, concretely. An open recall is a reason to check what the campaign says, not a reason to assume either the worst or the best. Nearly a quarter of car recalls are standards failures, and over half of the largest class of those is a display or software problem that a dealer visit resolves. But the second most-cited standard in this data is occupant crash protection - the airbag standard - so "noncompliance" is not a synonym for "minor". Read the campaign.

Method

Source. NHTSA Office of Defects Investigation Part 573 recall flat file, FLAT_RCL_POST_2010.txt, from https://static.nhtsa.gov/odi/ffdd/rcl/FLAT_RCL_POST_2010.zip, public domain. Downloaded 2026-10-21 07:19 UTC; server Last-Modified 2026-10-21 07:05:02 GMT; zip member timestamp 2026-10-21 07:02:12. ZipFile.testzip() returned None. 246,253 rows, 29 tab-delimited fields on every row. Field positions are those documented in RCL.txt (last updated May 2025).

Freshness asserted before computing anything. The newest Part 573 report received date (RCDATE, field 16) in the file is 2026-10-16, five days before the run, and the newest record creation date is 2026-10-19. A stale-but-parseable file is the known failure mode with NHTSA flat files, so the run aborts rather than annotates if the newest filing is more than 30 days old.

Universe. 246,253 rows collapse one-per-CAMPNO to 15,332 distinct campaigns, of which 13,701 carry RCLTYPECD='V'. "V" means vehicle, not car: it includes motorcycles, school and transit buses, heavy trucks, motorhomes and trailers. Campaigns were restricted to cars and light trucks with a two-part screen - a make must be a US passenger-car or light-truck brand, and for makes that also sell medium-duty, heavy-duty or two-wheel product the campaign is dropped if its model or component strings name a heavy or non-car vehicle. 5,294 campaigns pass, 8,223 are excluded, and 184 span both car and non-car makes and are excluded as mixed. Every campaign's class and the reason for it ship in the frozen CSV so the classification can be redone or rejected.

Noncompliance definition. A campaign is counted as a noncompliance recall when field 19 (FMVSS) carries a standard number. This proxy was validated rather than assumed. Compliance language - "does not comply", "noncompliance", "fails to meet the requirements" and similar - appears in the defect, remedy or notes text of 86.87% of campaigns that carry an FMVSS number and 0.59% of those that do not. The mismatch runs both ways and bounds the count: 64 campaigns covering 1,074,832 vehicles use compliance language with no standard number recorded, so 1,269 is a floor; and 13.13% of FMVSS-citing campaigns do not use the language, which the regex may simply be missing.

Field integrity. All 43 standard values are exactly three characters and all numeric, so the May 2025 reduction of field 19 to char(3) has not truncated anything in this set. No campaign cites more than one standard, so the per-standard columns do not double-count.

Aggregation. Campaign counts are distinct CAMPNO. Vehicle counts are POTAFF (field 12), taken as the maximum across a campaign's rows. Filer figures are grouped on a normalised MFGNAME (field 8), because NHTSA carries several spellings of one company; the raw spellings merged into each filer are listed in the CSV.

Reversal checks run, and the results either way. · (a) The restriction check, which was the live threat: on the unrestricted "vehicle" basis the answer is 2,802 campaigns and 69,269,667 vehicles, and restricting to cars and light trucks cuts the campaign count by 54.7% while cutting vehicles by only 15.4% - so the two summary statistics do not degrade together and the campaign count is the fragile one. Every figure above is on the restricted basis. (b) Concentration, the check most likely to reverse the rear-visibility headline: it confirmed it instead. The largest FMVSS 111 campaign is 7.6% of that standard's vehicles and the top three are 21.4%, against a median campaign of 18,411 vehicles, so the 16,192,156 is broad-based. (c) The same check disqualified other standards from being ranked at all: for 25 of the 43, one campaign is more than half the standard's vehicle total, FMVSS 105 being 99.2% a single campaign (24V051000, 2,193,869 vehicles) against a median of 88. The CSV carries a rankable flag per standard and this article ranks none of the disqualified ones. (d) A trend claim was computed and then refused: the share of car campaigns citing a standard rises from 14.29% in 2010 to 32.44% in 2026, which cannot be separated from improving completeness of NHTSA's own field, so no rising-compliance-recall claim is made here.

Known limits. · POTAFF is the manufacturer's launch-time estimate of units potentially affected. It is not a count of cars still on the road, still unrepaired, or still in the United States, so this file supports "how much exposure was opened" and cannot support "how many non-compliant cars are out there". The car and light-truck screen is ours and is a judgement, which is why it ships in the CSV. 2026 is a partial year. We did not contact NHTSA or any manufacturer before publishing. Nothing here measures how dangerous a given standard's failure is; campaign counts and unit counts are exposure, not harm.

Frozen dataset. All 43 standards with campaign counts, vehicle totals, per-campaign medians, the largest campaign and its share, and the rankability flag: car-recalls-failed-federal-standard-not-broken-part-nhtsa-data-2026-10-21.csv

Sources

• NHTSA Office of Defects Investigation, Part 573 recall flat file (public domain): https://static.nhtsa.gov/odi/ffdd/rcl/FLAT\_RCL\_POST\_2010.zip - retrieved 2026-10-21 07:19 UTC, Last-Modified 2026-10-21 07:05:02 GMT. Last verified 2026-10-21.
• NHTSA Office of Defects Investigation, recall file layout and change log: https://static.nhtsa.gov/odi/ffdd/rcl/RCL.txt - last updated May 2025. Last verified 2026-10-21.
• Electronic Code of Federal Regulations, 49 CFR Part 571, Federal Motor Vehicle Safety Standards. Standard titles were read from the eCFR API rather than the web page, which is bot-gated to automated fetch from our environment: https://www.ecfr.gov/api/versioner/v1/structure/2026-01-01/title-49.json (title-49 structure, 2026-01-01 edition), retrieved 2026-10-21, HTTP 200. Human-readable equivalent: ecfr.gov/current/title-49/part-571. Last verified 2026-10-21.
• NHTSA recall and VIN lookup, for remedy status (cited by address; not machine-retrieved for this analysis): https://www.nhtsa.gov/recalls - Last verified 2026-10-21.
• Zilocar frozen dataset for this analysis, linked in the Method section above.