⌘K
Browse the recordHome›Reports› Q3 2026

56 recalls filed against the 2026 model year in Q3 2026.

State of the model year · Q3 2026

State of the 2026 model year

A dated snapshot of the 2026 model year's federal recall record, built from NHTSA's own filings. as of 2026-09-30 — the figures on this page will not change. Every quarter →

Frozen · figures as they stood on 2026-09-30

Photo: kimak kimbek / Unsplash

In one sentence each

84%

clean or minor record

253 of 302 2026 model years on file are graded A or B on their NHTSA recall record.

5

graded F

Model years carrying the worst recall record of the year — the most popular: Ford Explorer, Ford Bronco.

14

under an advisory

Park-outside or do-not-drive, the agency's own words, on 2026 model years.

56

campaigns filed this quarter

Against the 2026 model year in Q3 2026, covering 6,482,010 vehicles by the manufacturers' own counts.

How the 2026 model year grades

Source: NHTSA recall record

302 model years on file. 23 are provisional — nothing has been filed against a model year that new, so the grade is scored on an empty record and should be read as untested, not as proven.

GradeModel yearsShare
A20167%
B5217%
C3512%
D93%
F52%

The ten most-recalled 2026 model years

Source: NHTSA

Under an advisory

Source: NHTSA

The agency is telling owners to change what they do with the car today — not at the next service. 14 2026 model years.

VehicleAdvisoryGrade
2026 Cadillac Escaladedo not driveD
2026 Chevrolet Silveradodo not driveD
2026 Chevrolet Suburbando not driveC
2026 Chevrolet Tahoedo not driveC
2026 Ford Bronco Sportdo not driveF
2026 Ford Maverickdo not driveF
2026 GMC Sierrado not driveD
2026 GMC Yukondo not driveC
2026 Volkswagen Atlasdo not driveB
2026 Volkswagen ID.4do not driveB
2026 Jeep Grand Cherokeepark outsideD
2026 Lucid Airpark outsideC
2026 Nissan Leafpark outsideC
2026 Subaru Crosstrekpark outsideB

Filed in Q3 2026

Source: NHTSA 573 reports, 2026-07-01 to 2026-10-01

Campaigns filed in the quarter that name the 2026 model year, by who filed them. Vehicle counts are the manufacturer's own estimate on the defect report.

ManufacturerCampaignsVehicles
Ford Motor Company161,517,972
Chrysler (FCA US, LLC)102,717,883
Toyota Motor Engineering & Manufacturing6523,442
General Motors, LLC4336,821
Nissan North America, Inc2171,937
Rivian Automotive, LLC2102,657
Hyundai Motor America253,147
BMW of North America, LLC227,733
Jaguar Land Rover North America, LLC21,074
Tesla, Inc234

By make

Source: the fifteen with the most 2026 model years on file
MakeModel yearsCampaignsGraded FAdvisories
Audi201200
BMW192100
Toyota191900
Ford168252
Mercedes-Benz161600
Chevrolet133003
Hyundai112700
Nissan111101
Lexus111000
Volvo11800
Honda10700
Subaru10701
Kia9900
Jeep82301
GMC81802

What a grade is, and is not

Source: method

A model year is graded on its NHTSA recall record: the grade starts at 100 and subtracts per campaign by severity; A-B is a clean or minor record. 'Provisional' means nothing has been filed against a model year that new, so it scores full marks on an empty record and should be read as untested. Advisories are the agency's own park-outside / do-not-drive flags. Campaign counts are campaign-level: a campaign names model years and build windows, never individual vehicles.

Sources: NHTSA recall campaigns (573 defect reports); NHTSA park-outside and do-not-drive advisories; EPA fueleconomy.gov (the nameplate list). Every figure on this page can be re-derived from the public record and from the JSON and CSV below.

The data behind this

Released under CC BY 4.0 — copy it, change it, build a product on it, including commercially. Credit WhichTrim and link back. See the carve-out in the terms.

Everything above, as files. No key, no form, no attribution beyond the line below — if it is useful, use it.

How to cite

WhichTrim, “State of the 2026 model year — Q3 2026”, 2026-11-04. https://whichtrim.com/reports/2026-q3/

The underlying records are U.S. federal public-domain data from NHTSA. The joins, the filtering and the arithmetic are ours, and we would rather be credited than not — but the figures are checkable from the source either way, which is the point.