Automobile Manufacturers Data: Registrations, Production, Powertrain and the Vehicle Registry · Head-to-head
EPA Fuel Economy Dataset (1984-2026) vs Washington State Electric Vehicle Population Data
Which automobile manufacturers data: registrations, production, powertrain and the vehicle registry data fits your job: EPA Fuel Economy Dataset, or Washington State Electric Vehicle Population Data. API, files, or your warehouse. Daily, weekly, or hourly.
EPA Fuel Economy Dataset (1984-2026)
Washington State Electric Vehicle Population Data
Coverage, side by side
| EPA Fuel Economy Dataset | Washington State Electric Vehicle Population Data | |
|---|---|---|
| Geographic | United States - every configuration certified for the US market; 49-state and California formulations distinguished in historical files | Washington State only, resolved to county, city, ZIP, legislative district, 2020 census tract and geocoded point, plus utility territory |
| Temporal | Model years 1984-2026 combined; per-year packages back to 1978 (2027 preliminary) | Current registered fleet, refreshed weekly; first published in this form October 19, 2023, last modified August 13, 2026 |
What each contains
They tie on 1 attribute. Pick by fit, not by loyalty.
| EPA Fuel Economy Dataset | Washington State Electric Vehicle Population Data | |
|---|---|---|
| Publisher | US EPA and US DOE, administered by Oak Ridge National Laboratory; tested at the EPA National Vehicle and Fuel Emissions Laboratory, Ann Arbor | — |
| What a row is | One vehicle configuration: model year x make x model x engine/transmission option, with its tested ratings | One registered vehicle: a BEV or PHEV titled in Washington State, located to census tract |
| Field dictionary | 23 documented fields, verified during research | 16 documented fields, verified during research |
| Geographic coverage | United States - every configuration certified for the US market; 49-state and California formulations distinguished in historical files | Washington State only, resolved to county, city, ZIP, legislative district, 2020 census tract and geocoded point, plus utility territory |
| Temporal coverage | Model years 1984-2026 combined; per-year packages back to 1978 (2027 preliminary) | Current registered fleet, refreshed weekly; first published in this form October 19, 2023, last modified August 13, 2026 |
| Scale | Tens of thousands of configuration records across 43 model years | 294,193 vehicle records as of August 2026 |
| Powertrain scope | All six fuel types: gasoline, diesel, CNG, ethanol, hybrid-electric and electric (charge time, MPGe) | Battery electric and plug-in hybrid only |
| Measurement basis | Laboratory testing plus manufacturer results under EPA oversight; revisions documented by manufacturer | Registration records; electric range as reported to DOL |
| Best for | How vehicles perform: tested efficiency, CO2 and specs across four decades | Who actually bought in: live EV brand share and neighborhood geography |
What each does better
the EPA Fuel Economy Dataset
Depth of time. The combined file spans model years 1984 through 2026 and per-model-year packages reach back to 1978 (2027 marked preliminary) - 43 years of tested efficiency in one consistent frame, against a registry that snapshots only the currently registered fleet and first went online in this form in October 2023.
Engineering-grade measurement. Every figure originates in testing at the EPA's Ann Arbor laboratory or from manufacturer results performed under EPA oversight, and the methodology is documented enough to cite: estimates for 1984-2007 and some 2011-2016 vehicles were revised for cross-year comparability, with revised values flagged for Nissan, Volkswagen Group, Hyundai/Kia and Ford. A registration's range figure is whatever the owner's paperwork said.
The whole powertrain landscape, not just the electrified edge. All six fuel codes appear - regular, premium, diesel, CNG, ethanol and electric, with EV-specific attributes such as charge time and MPGe riding along - so hybrids, diesels and V8s sit in the same table as Teslas. The registry, by design, contains only BEVs and PHEVs.
Configuration-level granularity. One row per year x make x model x engine/transmission option, plus passenger and luggage volume fields (2pv, 4pv, hpv, hlv) and class codes for model years 2000 onward - spec-sheet data no registry of titled vehicles attempts.
the Washington State EV Population Data
It counts real adoption, not rated capability. 294,193 registrations is the market that actually happened: which makes people bought, in what volume, and where they plugged them in. Tesla dominates the observed rows, with Nissan Leaf and Chevrolet Bolt/Volt present - brand-share evidence no laboratory dataset can produce, because laboratories never sell cars.
Geography to the neighborhood. County, city, ZIP, legislative district, 2020 census tract and a geocoded point per row, with electric_utility attaching each vehicle to its service territory (PUGET SOUND ENERGY INC and PACIFICORP both appear in the samples). Utility-territory load planning starts exactly here; see census tract geography. The EPA file cannot place a single vehicle below the national level.
The EPA combined file is rebuilt on an annual rhythm, last updated August 7, 2026.
An identifier that joins to the outside world. vin_1_10 preserves the first ten VIN characters - model year, plant and attributes encoded - truncated for privacy but still decodable to manufacturer and model line; see VIN decoding. The EPA dictionary keys on names alone.
Where they're equivalent
More than their shapes suggest. Both field dictionaries were verified during research, part of the 85.7 percent of cataloged records with confirmed documentation, and both score 9 of 10 - a band only about 39 percent of all 1,744 cataloged datasets reach. Both publish structured tabular records rather than narrative reports, both key every row to a named make and model, and both describe the United States rather than a global panel - the EPA file because it certifies US-market vehicles, the registry because it counts Washington plates.
They share their limits symmetrically too. Neither carries a historical series of market outcomes: the EPA file is timeless configurations without sales volumes, the registry is current population without the departed - so trend work on either side needs repeated sampling or outside volume data. And neither documents individual owners beyond what privacy truncation allows.
The verdict
Verdict: sample both, pick by fit - they are different instruments pointed at the same industry.
Take the EPA Fuel Economy Dataset if your question asks how vehicles perform. Efficiency benchmarking and CO2 modeling across four decades of powertrains, spec-level competitive teardowns by engine/transmission option, compliance and technology-positioning evidence, journalism citing official MPG figures under a documented revision methodology. Accept its frame: configurations, not customers - no geography below the nation and no sales counts.
Take the Washington State EV Population Data if your question asks who actually bought in. Brand share off live registrations, adoption mapping by census tract and legislative district, utility-territory load forecasting, charging-infrastructure siting. Accept its frame: one state, electrified vehicles only, and a range figure that is owner-reported rather than tested.
Data scientists usually want both - tested ratings to explain performance, registration reality to weight it; market researchers tend to start with the registry and reach for the EPA file when the question turns to product.
Sample both, pick by fit. See EPA Fuel Economy Dataset · See Washington State Electric Vehicle Population Data
Or take both in one feed
Yes - they stack into one picture neither completes, joined on the four concepts their dictionaries genuinely share: model year, make, model and powertrain. An analyst can show not just that Model 3s cluster around Seattle but what efficiency class and fuel bill those buyers signed up for.
Two alignments decide whether the merge holds. First, name normalization: manufacturer and carline name are written differently from make and model (TESLA versus Tesla, MODEL 3 versus Model 3), and the registry's vin_1_10 prefix resolves to model lines the EPA file keys on - normalize before joining, and use VIN decoding when ten characters need to become a maker and model. Second, period semantics: model years are not calendar years, so align a 2020-model-year configuration against registrations from its actual sales window before reading any trend off the pair. Browse the rest of the shelf at the automobile manufacturers data hub.
Datadory ships either record alone or both merged onto one vehicle list, delivered daily, weekly, or hourly - your call. Or take both in one feed.
API, files, or your warehouse. Daily, weekly, or hourly.
Fair questions
Is the EPA Fuel Economy Dataset better than the Washington State EV Population Data?
Better at different jobs. The EPA dataset owns measurement: 23 documented fields of tested city/highway/combined MPG - adjusted and unadjusted - plus CO2, displacement, cylinders, drive and fuel type, covering every US configuration across model years 1984-2026 with a documented revision methodology. The Washington registry owns reality: 294,193 actual BEV and PHEV registrations with county, city, ZIP, legislative district, census tract, utility territory and electric range, refreshed weekly. Sample both, pick by fit.
Do the two datasets cover the same ground?
Only four concepts deep: model year, make, model and a powertrain label exist on both sides, written in different vocabularies. Everything else diverges by design - the EPA file spends its fields on laboratory measurement and configuration detail, Washington spends its sixteen on registration geography down to census tract and utility territory. One describes how every US vehicle performs; the other records which electrified ones Washingtonians actually drive.
Can I join EPA fuel economy ratings to EV registrations?
Yes - join on model year, make, model and powertrain after normalizing names: the EPA file writes `manufacturer` and `carline name` while the registry prints `make` and `model` in uppercase, so TESLA must match Tesla before anything merges. Two caveats from the records: Washington truncates VINs to ten characters, and model years are not calendar years - align a configuration to its sales window, not its badge year. Datadory handles both normalizations in a merged feed.
Which dataset shows real-world EV adoption?
The Washington State registry, unambiguously. It counts 294,193 registered BEVs and PHEVs as of August 2026 - who bought what, refreshed weekly - with Tesla dominating observed rows and Nissan Leaf and Chevrolet Bolt/Volt among tagged makes, resolvable to county, city, ZIP, legislative district and census tract. The EPA file has no buyers in it at all: it documents configurations and their tested ratings, never volumes sold.
Does the EPA dataset include electric vehicles too?
Yes. Electric is fuel-type code El among six - R regular, P premium, D diesel, C natural gas, E ethanol - and EV rows carry dedicated attributes such as charge time and MPGe alongside the standard MPG, CO2 and engine fields. That full-powertrain span is precisely what the Washington registry lacks, since it holds only BEVs and PHEVs; pairing them puts electrified registrations inside the whole market's performance frame.
Can Datadory deliver both datasets together?
Yes - alone or merged onto one vehicle list, delivered daily, weekly, or hourly, your call. Each arrives normalized to its documented dictionary (23 fields on the EPA side, 16 on Washington's) with sample rows for inspection before anything ships. The joining work is name normalization across make, model and model year, which we handle in the merge. Or take both in one feed.