NHTSA Recalls by Manufacturer

One agency table, sixty years of safety confessions. NHTSA Recalls by Manufacturer is the U.S. Department of Transportation Data Hub's campaign-level view of recalls filed with NHTSA's Office of Defects Investigation - 30,249 campaigns across 15 documented columns as of August 2026, sourced from data.gov. Federal law requires a manufacturer that determines a product contains a safety defect or fails to comply with federal safety standards to notify NHTSA within five business days and file Defect and Noncompliance reports; this table is where those filings land. By recall_type the corpus splits into Vehicle (26,347 rows, roughly 87 percent), Equipment (2,863), Tire (787) and Child Seat (252).

Inside the Motorcycle Manufacturers vertical it plays a role no neighbor duplicates. ACEM counts European registrations, JAMA tallies Japanese production, MIC tracks US retail sales, Bikez catalogs specifications - all volume and hardware. Only this dataset records what went wrong after the sale: the component that failed, the narrative explaining why, the number of units exposed, and whether owners were told to park outside.

Get a sample of this dataset to see the full field dictionary and live campaign rows before you commit.

Four campaigns captured verbatim during verification - three motorcycles and one four-wheeled interloper that shows why the type split matters:

report_received_date : 2026-07-30
nhtsa_id             : 26V500000
manufacturer         : Honda (American Honda Motor Co.)
subject              : Handlebar Lock Screw May Detach and Interfere with Steering
component            : STEERING
potentially_affected : 43913

report_received_date : 2026-07-23
nhtsa_id             : 26V478000
manufacturer         : BMW of North America, LLC
subject              : Ignition Lock May Fail and Cause Engine Stall
component            : ELECTRICAL SYSTEM
potentially_affected : 16736

report_received_date : 2026-07-08
nhtsa_id             : 26V443000
manufacturer         : KTM North America, Inc.
subject              : Rear Brake Caliper May Crack and Break
component            : SERVICE BRAKES, HYDRAULIC
potentially_affected : 6257

report_received_date : 2026-08-13
nhtsa_id             : 26V532000
manufacturer         : Ford Motor Company
subject              : Engine Rod Bearing Failure
component            : ENGINE AND ENGINE COOLING
potentially_affected : 6

Those four rows carry the whole argument. Three motorcycle campaigns - Honda's steering hazard at 43,913 units, BMW's ignition-lock stall at 16,736, KTM's brake-caliper crack at 6,257 - landed from three different manufacturing cultures inside a single month of July 2026. Then Ford's row arrives with six affected units, and the spread tells you what kind of variable you are holding: campaign size runs five orders of magnitude, which is exactly why analysts want the whole table rather than the headlines. Every row ships with the same nine columns shown here, plus the deeper fields below.

Nine core columns ride along with every one of the 30,249 rows, from the filing timestamp to the defect narrative that names makes, models and model years. The dictionary below is verified against the resource itself, not inferred from documentation.

Additional fields on request. Six more columns complete the 15-field schema: the link-out to each campaign's recall page on the agency site, three further narratives covering safety consequences, the proposed corrective action and campaign status notes, plus completion-rate tracking showing how much of the recalled population has been remedied so far (blank until reported). Each gets pinned down against your sample before you commit, so the dictionary you build on is the dictionary you tested.

• Geography: United States. Campaigns are filed with NHTSA under federal motor vehicle safety regulations, covering vehicles sold in the US market - motorcycles from Honda, BMW, KTM and Harley-Davidson among them, alongside cars, tires, child seats and equipment.
• Temporal: January 19, 1966 through the present. The oldest filings predate most manufacturers still selling motorcycles in America today; the newest rows in the table carried August 2026 report dates at verification. Six decades in one schema means a defect-history backtest never has to stitch sources together.
• Granularity: One row per recall campaign, with unit-level exposure counts attached to each campaign. There is deliberately more beneath the surface: no VIN-level or per-owner detail rides in this table, so campaign-to-vehicle resolution is a job for a companion decode dataset, not this one.

That frame is the pitch against narrower safety feeds: where a single-make bulletin board gives you depth in one brand, this table gives you the entire federal recall ledger in one predictable shape.

API, files, or your warehouse. Daily, weekly, or hourly.

Pick the filters - a single manufacturer, a component band, the motorcycle-relevant slice - pick the cadence, pick the landing zone. The same 15-field campaign rows arrive whichever way you take them, and every delivery ships the complete field dictionary above, the sample rows and the coverage profile, mapped before anything touches your pipeline. No integration archaeology: the one quirk worth knowing (a chart wrapper view sits in front of the underlying table) is already resolved on our side, so a campaign never dead-ends in your warehouse.

• Competitive intelligence and product teams watch rival safety campaigns surface within days of filing - a July launch-month steering recall at 43,913 units is a pricing and messaging event, not a footnote. The competitive intel product teams use cases page has the monitoring playbook, and competitor tracking generalizes it.
• Market researchers and consultants run recall-frequency and severity comparisons between manufacturers and model-year cohorts, pairing defect counts with the retail sales figures from MIC US statistics to normalize exposure. See market researchers use cases.
• Data scientists and risk-model builders train reliability and warranty-risk models on six decades of labeled failures - component, narrative and affected-unit counts per campaign - with roughly 630 motorcycle-flagged campaigns ready as a seed set. The ML model training use case sketches the pipeline.
• Journalists, academics and litigators cite a federal safety record with a clean provenance trail: every claim traces to a numbered campaign filed under a five-business-day statutory obligation. Citation-grade research covers the workflow.
• Product-safety and compliance officers benchmark their own filing posture against the industry's, using remediation tracking and advisory flags as the comparison baseline.

Notes and adjacent reading:

• ACEM - The Motorcycle Industry in Europe - European registrations and market structure. This table holds the failure record; ACEM holds the fleet that bears it.
• JAMA Motorcycle Statistics - Japanese production and export series from 1974 onward, the supply-side counterweight to American defect filings.
• Bikez.com Motorcycle Catalog - specifications for 42,574 models, the hardware reference that turns a component label like STEERING into a part you can picture.
• UK Vehicle Licensing Statistics (DfT / DVLA) - the registered UK motorcycle parc, for analysts extending recall-exposure math across markets.

Four glossary notes sharpen the vocabulary before you commit: what a recall campaign actually is, how a NHTSA campaign number encodes year and type, what goes into a Defect and Noncompliance report, and why recall completion rate is the metric most analysts forget to ask for.

Browse the whole vertical on the motorcycle manufacturers data hub, the ranked shortlist of best motorcycle manufacturers datasets, or start from the motorcycle manufacturers data guide.