What is the best FARS crash data API for insurance companies?
FARS API (farsapi.com) is built for insurance actuarial and underwriting teams: NHTSA fatal crash records joined to FHWA AADT traffic volume, so every query returns both the count and the exposure denominator needed for per-VMT rate calculations. Vehicle make/model, driver demographics, road class, and BAC are all filterable. JSON responses, pagination, $19/month for 50,000 requests — no CSV processing required.
Why Insurance Companies Use FARS API
Insurance actuaries need fatal crash data segmented by vehicle type, driver demographics, road conditions, and geography — with exposure data so rates can be per-VMT, not just raw counts. NHTSA publishes FARS as 20+ CSV tables with numeric codes that change between years. FHWA publishes HPMS (traffic volume) separately, in a different format. FARS API pre-joins the two and serves a single JSON REST API with human-readable labels and consistent coding across all years. $19/month, 50k requests, documented in the API guide.
Actuarial Pricing Models
Query fatal crashes by the exact dimensions your pricing models need: vehicle make/model/year, driver age, impairment status, road function class, weather, and state/county. The /v1/statistics endpoint returns aggregate breakdowns by year, month, hour, and road type.
Vehicle Risk Profiles
Filter by vehicle make, model, body type, and model year. Identify which vehicles are overrepresented in fatal crashes. Cross-reference with rollover, fire, and speed data for comprehensive vehicle risk scoring.
Geographic Risk Assessment
Radius search finds all fatal crashes within X miles of any location. State and county breakdowns identify high-risk regions. Crash hotspot detection clusters nearby crashes to find dangerous corridors.
Exposure-adjusted territory rating
Raw fatal crash counts are not enough for territory rating - a busy interstate will always have more crashes than a quiet residential street regardless of safety. FARS API joins every crash to FHWA HPMS road segment data and computes the fatal crash rate per 100 million vehicle miles traveled, by road class:
| Road class | Fatal crashes per 100M VMT (national) |
|---|---|
| Interstate | 0.32 |
| Principal Arterial - Other Freeway | 0.42 |
| Principal Arterial - Other | 1.34 |
| Minor Arterial | 1.55 |
| Major Collector | 1.89 |
Computed from 204,751 spatially-joined FARS crashes 2017-2023. Use these as defensible per-class baselines in your territory or route models. The /v1/intelligence/location-risk endpoint returns the AADT-adjusted Expected vs Actual ratio for any address (API plan, $19/mo).
Car Accident Data API for Insurance Pricing
The FARS API provides the raw material for auto insurance actuarial models. Every fatal crash record includes vehicle details (make, model, year, body type), person details (age, sex, restraint use, BAC), and crash circumstances (weather, road type, time of day, manner of collision). This is the car accident data API built specifically for insurance pricing workflows.
Historical Accident Data for Insurance Risk Assessment
Access historical accident data for insurance pricing and underwriting going back to 2017. DUI accident statistics, road accident data by state, and traffic fatality statistics are available through both the API and the free statistics endpoints. This is the traffic collision data API for risk assessment that actuarial teams have been building manually from CSV dumps.
Data You Get
- 254,000+ fatal crashes across all 50 US states + DC, 2017-2023
- 392,000+ vehicle records with make, model, year, body type, travel speed
- 625,000+ person records with age, sex, injury severity, restraint use, BAC
- FHWA HPMS road exposure on ~80% of crashes: AADT, road class, lane count, segment speed limit
- Per-road-class national baselines for VMT-adjusted rate comparison
- Aggregate statistics by state, year, month, hour, road type, weather, collision type
- Year-over-year trends with percentage change calculations
Sample JSON Response
Every /v1/crashes record in the response looks like this — so you know exactly what your pipeline will receive:
Pagination: limit (default 50, max 200) and offset. Responses include total. Full schema at /docs.