One in Six US Car Recalls Is Filed Under More Than One Component — and NHTSA's Own Lookup Shows You Only One

Key facts

• 13,649 distinct US vehicle recall campaigns (RCLTYPECD = 'V') in NHTSA's post-2010 recalls flat file, retrieved 2026-10-01.
• 2,144 campaigns (15.71%) carry more than one distinct COMPNAME component label. Restricted to cars and light trucks only, 1,607 of 10,200 — 15.75%.
• Label distribution: 11,505 campaigns with one label, 1,951 with two, 164 with three, 26 with four, 2 with five, 1 with six.
• Multi-label campaigns account for 104,211,137 of 536,874,791 potentially affected units — 19.4% of all recall exposure opened since 2010.
• In 12 of 12 multi-label campaigns we queried, api.nhtsa.gov/recalls/campaignNumber returned exactly one component label.
• Worked example: Honda campaign 20V771000 (2018-2020 Accord, Accord Hybrid, 2019-2020 Insight; 737,233 units) carries six labels in the flat file. The API returns one.
• Live example: Cadillac 26V612000, filed 2026-09-24 (2025-2027 Optiq, 29,347 units), is filed under both power window controls and body control module software.
• A VIN check confirms a recall exists and how many are on file. It does not confirm a repair was performed.

Why does one car recall get more than one component label?

Usually because a single defect has its cause in one system and its symptom in another, and NHTSA files it under both. The clearest live case is Cadillac's 26V612000, filed 2026-09-24 for 29,347 Optiq SUVs from model years 2025-2027. NHTSA's own defect description says that when the front windows are closed using the key fob's express-up function they "may fail to detect an obstruction and reverse direction," and that the remedy is a body control module software update. So the campaign is filed under two labels: VISIBILITY:POWER WINDOW DEVICES AND CONTROLS — what the owner experiences — and ELECTRICAL SYSTEM:BODY CONTROL MODULE:SOFTWARE — what is actually broken.

Neither label is wrong. Together they describe the recall properly. Separately, each one tells a partial story, and which of the two you see depends on which NHTSA surface you happen to read.

Does NHTSA's recall lookup show every component?

In our testing, no. We took twelve multi-label campaigns filed since 2025-01-01, chosen largest-first by units potentially affected, and queried api.nhtsa.gov/recalls/campaignNumber for each. In all twelve the API returned exactly one component string, even for campaigns the flat file records under four. This is the same API that powers public-facing recall lookups, so the omission travels.

The most striking case is not recent. Honda's 20V771000, filed 2020-12-10 for 737,233 Accord, Accord Hybrid and Insight cars, is filed in the flat file under six labels: back-over prevention camera, body control module software, exterior lighting, wipers, defroster/HVAC, and power window controls. The API returns only ELECTRICAL SYSTEM:BODY CONTROL MODULE:SOFTWARE.

What makes that a fair criticism rather than a technicality is that NHTSA's own consequence text for the same campaign reads: "Various system malfunctions such as inoperative windshield wipers, defroster, rearview camera, or exterior lighting can increase the risk of a crash." The prose names the systems. The structured component field, which is what tools and filters read, names one of them.

Which label survives is not predictable. On the Optiq the API kept the symptom (power windows) and dropped the cause (BCM software). On the Honda it kept the cause and dropped all five symptoms. On the Ford Ranger it kept BACK OVER PREVENTION:DISPLAY FUNCTION and dropped EQUIPMENT:ELECTRICAL:INFOTAINMENT. On GM's 2.4-million-car 14V252000 it kept ELECTRICAL SYSTEM:WIRING and dropped brake lights, the brake-light switch, electronic stability control and cruise control — four of the five labels, including the two that name the part a buyer would actually test.

Does a multi-label recall mean the defect is more serious?

No, and this is the most important thing not to conclude from these numbers. The label count measures how NHTSA's filing system categorised the paperwork, not how dangerous the defect is or how many cars are affected. A campaign with six labels and a campaign with one can carry identical risk. The Optiq recall above has two labels and a single, well-understood software fault with an over-the-air remedy.

This is the same caution that applies to counting complaints by category: a count of filing categories measures filing behaviour. We found the same thing when we audited whether do NHTSA complaint numbers mean a used car is unreliable — a large share of what looks like defect signal is correspondence about recalls, not reports of faults. Read the defect and consequence text; the structured fields are an index, not a verdict.

Has this become more common over time?

The measured share rises sharply and then settles. Campaigns filed in 2010 and 2011 are multi-label 1.5% of the time; by 2016 it is 15.0%, it peaks at 31.0% in 2021, and 2026 filings sit at 17.1%.

We do not read the early figure as evidence that recalls became six times more complex. A jump from 1.5% to 9.6% between 2011 and 2012, holding the rest of the file's structure constant, looks like a change in how NHTSA records components rather than a change in the recalls themselves. The honest statement is that multi-label filing has been normal practice for at least a decade and the recent run sits in the high teens to low thirties. Anyone wanting a trend claim should treat pre-2012 filings as not comparable.

What does this mean when you are buying a used car?

It means the component label on a recall is a starting point, not the scope. If you are checking a used car and a recall comes back tagged "electrical system," the campaign may also be filed under the wipers, the defroster and the rearview camera — as Honda's 20V771000 is — and the thing that will actually annoy or endanger you may be in the part you were not shown.

Two practical consequences:

1. Read the campaign, not the tag. Look up the campaign number on NHTSA's recall pages and read the defect description, the consequence and the remedy. Those are prose fields and they name every system involved.
2. Do not filter a car out, or in, on the component label alone. A buyer searching only for "brake" recalls will miss a campaign filed under a module whose consequence is brake-related.

If you are weighing whether an open recall should change your decision at all, that is a separate question from scope, and we have looked at is it safe to buy a used car with an open recall against NHTSA's completion data. A related trap is assuming a closed campaign means a fixed car: see does a completed recall mean a used car is fixed.

By the numbers

• 13,649 vehicle recall campaigns; 218,890 component rows.
• 2,144 multi-label campaigns (15.71% of all; 15.75% of cars and light trucks).
• 193 campaigns with three or more labels; 29 with four or more.
• 19.4% of all potentially-affected units sit in a multi-label campaign.
• 12 of 12 sampled multi-label campaigns returned one label from the NHTSA API.

Based on 13,649 US vehicle recall campaigns in the NHTSA ODI recalls flat file, retrieved 2026-10-01, with a newest filing date of 2026-09-29.

What a VIN check can and can't tell you here

A Zilocar VIN check can tell you that a specific vehicle has recalls on file and how many, using NHTSA's own records — which is the screening question that matters when you are standing in front of a car. It also surfaces the vehicle's accident and damage records, odometer and rollback checks, theft records, junk and salvage auction records, ownership history, sales-listing history and valuation, none of which NHTSA holds.

It cannot tell you the full component list for a campaign, whether a recall was remedied or the software flashed, or whether a vehicle falls inside an open NHTSA investigation (a PE or EA number, which is not a recall). For the component scope, the authority is the campaign's Part 573 report on NHTSA's site. For remedy status on one VIN, the authority is a franchised dealer for that brand or NHTSA's own VIN recall tool. We do not claim otherwise, and a tool that claims to confirm a repair from public data is claiming something the data does not contain. The difference between an open recall and an open investigation is covered in what does recall closed mean on a vin check.

Method

Source file. NHTSA ODI recalls flat file FLAT_RCL_POST_2010.txt, downloaded from static.nhtsa.gov/odi/ffdd/rcl/FLAT_RCL_POST_2010.zip on 2026-10-01 at 07:05 UTC. The archive reported Last-Modified: Thu, 01 Oct 2026 07:04:57 GMT; its single member is 311,299,755 bytes with an internal timestamp of 2026-10-01 07:01:58, and testzip() returned no error, so the archive is intact. The file is tab-delimited, 29 fields, no header row.

Freshness assert, run before counting anything. The newest RCDATE (the date the Part 573 report was filed) is 2026-09-29, two days before retrieval. NHTSA's combined complaint download has twice served years-stale extracts under a fresh timestamp, so this check runs first and the analysis aborts if the newest record is more than 30 days old.

Universe. 245,627 rows in total; 218,890 carry RCLTYPECD = 'V', grouping into 13,649 distinct campaign numbers. RCLTYPECD = 'V' means "vehicle", not "car": it also carries motorcycles, buses, heavy trucks, motorhomes and towable equipment. Campaigns whose every MAKETXT falls in an explicit non-car exclusion list are classified other_vehicle; the remaining 10,200 are car_light_truck. Both bases are published above, and the classification ships as a vehicle_class column in the frozen CSV so a reader can redo or reject it.

The count. For each campaign number, the distinct values of COMPNAME are collected and counted; a campaign with two or more is multi-label. Rows are one per component-per-vehicle, so a campaign appears many times — the count is of distinct labels per campaign, never of rows.

Result. 2,144 of 13,649 = 15.71%. Cars and light trucks only: 1,607 of 10,200 = 15.75%.

Reversal check 1 — is this an artifact of one campaign covering several different vehicles that each got a different part? No. For 2,144 of 2,144 multi-label campaigns (100.0%), at least one single make/model/model-year combination by itself carries more than one label. Zero campaigns are multi-label only across different vehicles. The finding is about one car, not a campaign's spread.

Reversal check 2 — does restricting to cars move the number? No. The share moves from 15.71% to 15.75%, and the unit-weighted share is 19.4%. Both summary statistics survive the restricted basis, which is not automatic: an earlier study's campaign count nearly halved under this same check while its unit count barely moved.

The API comparison. · api.nhtsa.gov/recalls/campaignNumber?campaignNumber=<id> returns one Component string per affected model-year record. For twelve multi-label campaigns filed since 2025-01-01, sampled largest-first by units, the API returned fewer labels than the flat file in 12 of 12, collapsing to exactly one each time. The named examples in the table above were each re-queried individually.

Known limits.

• Units potentially affected (POTAFF) is the manufacturer's launch-time estimate. It is not a count of cars still on the road, still unrepaired, or still in the United States, and nothing here supports a model-level "worst car" ranking.
• COMPNAME is a controlled but inconsistently punctuated label. Grouping at the top level before the first colon is used only for description, never for the headline count.
• Manufacturer-level figures group on MFGTXT, the Part 573 filer, not MAKETXT, because one campaign often spans several makes.
• The twelve-campaign API comparison is a sample, not a census of the API's behaviour.
• Pre-2012 filings are not comparable to later ones on this measure; see the trend section.
• NHTSA republishes this file continuously, so a later re-pull will not reproduce these counts exactly. The frozen CSV is the reproducible artefact.

Sources

• NHTSA ODI recalls flat file, FLAT_RCL_POST_2010.zip — primary dataset, retrieved 2026-10-01 07:05 UTC.
• NHTSA datasets and APIs — file layout and publication practice.
• NHTSA recall API, campaign 20V771000 — the single-label API response for the six-label Honda campaign, including the consequence text quoted above.
• NHTSA recalls portal — the authoritative record for a campaign's full component scope and Part 573 report.
• Frozen dataset for this article (CSV) — all 13,649 campaigns with their label counts, labels, vehicle class and units.

Last verified: 2026-10-01