How we count
California's vehicles
What does this site count? 30,118,651 vehicles in 1,800 California ZIPs, from the DMV file of December 31, 2024. Make counts are lower bounds.
A snapshot of California's vehicles
Our counts for every fuel type start with the California DMV's public vehicle fuel type counts by ZIP code. Data as of December 31, 2024, the snapshot date recorded in the file. The state's open data portal lists the dataset's rights as "No Restrictions on Public Use".
These are aggregated vehicle counts, grouped by ZIP, make, fuel, model-year band and reported duty class. They are not individual vehicle records.
View the DMV datasetHow do the counts become claims?
Printed claims follow fixed rules, so a surprising number has to pass the same test as a dull one.
- Floors: the DMV pools a make's name where its count in one ZIP, model year and fuel group is below ten. A make count is therefore a lower bound. It prints as "at least" or with a plus sign, and every printed lower bound rounds down.
- Masking bias: pooled vehicles are a larger share of small ZIPs. Across the 1,119 ranked ZIPs, split into ten size groups, the pooled share runs from 31.7% in the smallest group to 5.7% in the largest. Across all 1,800 ZIPs it runs from 85.0% to 6.0%. Shares of named makes read low in small ZIPs. We flag this effect and do not correct it.
- Rankings: "highest" and "first" use only the 1,119 ZIPs with 5,000+ residents and 2,000+ vehicles, fleet ZIPs removed.
- Fleet rule: a whole ZIP leaves rankings above 130 vehicles per 100 residents with more than 35% from model years 2023 to 2025, or at 1.25 times growth in one year with 1,000+ vehicles. The file names no fleets, so this is a test on the numbers. It is not a record of who holds the vehicles. Fuel rankings also subtract a whole outlier make, or the model-year excess of a block, from the numerator and the denominator.
- Leads: a make leads a place only when its margin over the next named make exceeds the vehicles the DMV pools there. Otherwise the page calls it the largest make the DMV names.
- Small ZIPs: below 1,000 residents or 1,000 vehicles, per-resident ratios, household shares and percentiles do not print, and search engines are asked not to index the page.
- County rates: a rate per 100,000 residents for a county, such as a Cybertruck rate, adds up each ZIP's residents under its first-listed county. It is an estimate for ZIPs that cross a county line.
What do these words mean here?
- Registration
- A vehicle a state file counts at a place on its snapshot date. It is not a sale, and it does not say who owns the vehicle or where it is driven.
- Named make
- A make the DMV prints by name in a row of its file.
- Pooled make
- A vehicle the DMV counts under Other/Unknown, because its make group is too small to name in that row.
- Plug-in
- A battery electric vehicle or a plug-in hybrid.
- Zero-emission
- The California Energy Commission's group: battery electric, plug-in hybrid and hydrogen fuel cell vehicles. It includes plug-in hybrids, so it is larger than the battery electric count.
- ZIP and ZCTA
- A ZIP is a postal code. A ZCTA is the Census area that approximates it. Our map and population figures use ZCTAs, and our pages call both ZIPs.
A disclosed count is not always the full count
The DMV pools small make groups into Other/Unknown. That means a published count for a rare make can be incomplete. An absent make is not proof that no vehicles of that make exist in an area.
We keep those limits visible, especially for supercars and luxury makes. The public all-fuel file does not provide model names, license plates, owners or finance payments.
Make names need context
The source mixes cars, trucks, buses and specialty vehicles. White is a truck maker, not a color.
We merge labels that name the same maker: Blue Bird and Bluebird, IC Bus and IC Bus LLC, Thomas Built Buses and Thomas Bus, GMC and White/GMC, and Kovatch and the misspelled Kovatch Moblie Equipment Corp. The source's 124 named labels become 116 makes. Totals do not change. Each make page lists its source labels.
Charger, North American and Trolley name no maker. They join Other/Unknown, with the makes the DMV pools itself. Charger is not assumed to mean Dodge.
BYD, MG and MINI keep their spelling. We do not merge makes that share a parent company. Pooling still limits every named-make count.
NHTSA manufacturer and make recordsA ZIP is a place to start
The map fills Census ZIP Code Tabulation Areas, or ZCTAs. Darker color indicates a larger count or a larger share of that ZIP's fleet. ZIP labels are matched to the 2024 Census representative-point join inside California, then to 2020 Census cartographic area shapes.
More detail
All 1,800 mapped ZIPs have matching area geometry. ZCTAs approximate postal ZIPs; they are not city boundaries or vehicle locations. Large rural areas can visually dominate a map, so compare local shares and use the ranked tables as well as the colors. Color breaks are recalculated for the current filter selection.
Search covers 1,618 California incorporated places and Census-designated communities. A city or community search centers the map without changing the statewide summary or rankings. Vehicle filters apply across all mapped California ZIPs. Click a ZIP area to open its profile with those filters. These figures are not exact city totals. Hold Command while scrolling on Mac, or Ctrl on other computers, to zoom the map.
Every ZIP has a city or postal-locality name from GeoNames, used under CC BY 4.0. These mailing names are labels, not municipal boundaries. County overlaps come from the 2020 Census ZCTA–county relationship file. A “+ county” label means the ZIP area crosses a county boundary; its profile lists every county. The main county uses a matching California postal county where available, otherwise the largest California land overlap. Vehicle counts remain assigned to the whole ZIP. County pages count each ZIP once, in its main county, so 169 ZIPs that cross a county line count in one county only. City pages group ZIPs by postal city name, which does not match a city's municipal boundary.
People, homes and vehicle counts
The Census American Community Survey (ACS) 2020–2024 five-year release provides population, households and households with no vehicle available for all 1,800 mapped ZIP Code Tabulation Areas. These are estimates over a period, not an exact population count on the vehicle snapshot date.
More detail
The mapped areas sum to 39,282,836 estimated residents and 13,546,042 households. These are whole-ZCTA sums, including border areas, not an official California state population or city population. A Beverly Hills mailing name does not make ZIP 90210 equal to Beverly Hills city.
Vehicles per 100 residents means 100 times the selected vehicle count divided by the same area's estimated population. The denominator stays fixed when vehicle filters change. Registration counts include business and heavy-duty vehicles, so the rate can exceed 100 and is not a percentage of residents who own a vehicle.
Households with no vehicle is a separate Census measure of cars, vans and qualifying light trucks available for household use. It is not inferred from DMV registrations. Individual population and household margins of error use the Census 90% confidence level. Small-area rates can be unstable; zero denominators produce an unavailable rate. Per-resident rankings require at least 1,000 estimated residents. No margin of error has been calculated for the derived registration ratios.
All fuel types belong here
The all-fuel snapshot contains nine publisher categories. The charts group categories only where the label says so.
More detail
Share of ZIP fleet uses all reported counts in that ZIP as its denominator. A make's own profile uses its disclosed subtotal for the fuel and model-year mix. Neither is a share of residents or households. Duty class is the source's Light, Heavy or Unknown field, not a body-style classification such as SUV, sedan or coupe.
The pictures and share cards
Illustrations represent a make, model family or vehicle class. They do not identify a specific registered vehicle, generation, model year or trim. Heavy-duty selections use matching truck or commercial-vehicle artwork where available. Ambiguous labels and unsupported combinations use a coastal scene.
More detail
The DMV class field distinguishes light duty, heavy duty and unclassified vehicles. It does not give a sedan, SUV or coupe breakdown. A pictured SUV does not mean every vehicle of that make is an SUV.
Share cards highlight the count for the selected place, make or model. ZIP cards follow the active fuel, class and model-year filters. A card with no available count is not offered. Make counts retain DMV pooling limits; plug-in model cards keep their powertrain scope; new-retail cards count only new retail registrations, without fleets. The source dates, definitions and limitations on this page apply to the cards.
Model names come from a different source
The California Energy Commission publishes model-level counts for battery-electric, plug-in-hybrid and fuel-cell vehicles. This site's catalog retains 241 named make/model combinations from its December 31, 2025 California county snapshot. Matching make profiles show their model tables.
More detail
These tables cover plug-in and fuel-cell models only. They cannot provide conventional-fuel model totals, fill DMV's pooled make groups or be added to the December 2024 DMV counts. Model-year charts describe the mix of a snapshot, not sales or registration trends.
California Energy Commission (2025). Light-Duty Vehicle Population in California. Data last updated April 28, 2026. Retrieved November 6, 2026 from https://www.energy.ca.gov/zevstats.
The context around the cars
ZIP profiles and comparisons use ACS 2020–2024 tables B19013 (median household income), B19001 (household income distribution) and B25077 (median owner-occupied home value). All 1,800 map keys join to the source; usable income estimates exist for 1,589 and usable home-value estimates for 1,598. Unavailable estimates remain unavailable.
More detail
Income uses 2024 inflation-adjusted dollars. Census 90% margins of error appear with the estimates. The $200,000+ household share divides B19001's top income band by its household total; no margin of error is calculated for that derived share. Census open-ended medians display as $250,000+ or $2,000,000+, rather than as exact values. These estimates describe whole ZCTAs, not municipal populations or individual vehicle owners.
IRS Statistics of Income 2023 data adds rounded return counts, adjusted gross income bands, taxable interest, ordinary dividends and net capital gains. It matches 1,482 map ZIP labels. Unmatched ZIPs, including 92067, remain unavailable. Dollar amounts are converted from source thousands of dollars. Return counts are rounded to 10; small income cells can be combined or excluded, and dominant items can be suppressed. A combined band shows as Under 20, and an excluded item as Not published. Totals are reported aggregates, not complete local wealth.
IRS data uses return addresses, which can differ from residence. Returns were filed and processed in calendar 2024 and do not represent all residents. We do not join tax records to individual vehicles or divide returns by Census residents.
A reference for the machines
The model reference contains 50,407 FuelEconomy.gov configurations across 145 makes and 5,689 normalized model labels, covering published years 1984–2027. Fuel, transmission, drive, displacement, cylinder count, MPG/MPGe and available range are retained by model year. This is a U.S. specification catalog, not evidence that a configuration is registered in California.
More detail
Only whitespace and case variants are combined within model names. Distinct model labels remain separate, including body styles and regulatory names. Make aliases use an explicit crosswalk; McLaren Automotive becomes McLaren. No fuzzy model-family match assigns registration counts to specifications. Zero or blank ratings are not presented as usable range values.
The 2025 EPA test-car workbook adds exact normalized make/model label matches. Rated horsepower describes the regulatory engine/test entry; it is not assumed to equal combined hybrid system output. Repeated test-cycle entries with the same displayed mechanical fields are grouped. These entries are test records, not counts of cars. The workbook's equivalent test weight is not used as curb weight.
Five selected manufacturer notes preserve variant and original units. Their source dates: Porsche 2025 U.S. 911 Carrera, May 28, 2024; Lamborghini Revuelto, Page checked November 3, 2026; Aston Martin Vantage coupe launch specification, February 12, 2024; Rolls-Royce Spectre launch specification, July 4, 2023; Bentley Fourth-generation Continental GT Speed coupe, June 25, 2024. The 2025 Porsche 911 Carrera MSRP is a dated U.S. launch price with exclusions, not a current used-car valuation. Other models have no inferred price. Manufacturer photos and marketing descriptions are not reproduced.
FuelEconomy.gov's published terms permit noncommercial scientific and educational use. This research build uses that scope; unrestricted commercial redistribution has not been established. Free access to a source does not itself grant broader reuse rights.
The “Luxury & exotic” filter is an editorial make selection, including premium brands and exotic specialists. It does not classify the price or trim of individual cars. The all-vehicle map and unfiltered reference include all published makes.
What is arriving now?
The new-registration view uses the public Q3 2026 California Auto Outlook, released October 15, 2026, with data through September 30, 2026. It includes 36 named brands and Other, with Q3 and January–September comparisons, labeled latest quarter and year to date elsewhere on the site. Brand counts include fleets; the top five models in each of 18 segments, 90 rows, cover new retail registrations and exclude fleets. Percentages retain the report's one-decimal rounding. Absent brands and models are not assumed to have zero registrations.
More detail
The report's revised 2025 comparisons are retained together with its publication vintage. Registrations follow state title processing and can lag sale dates. These period flows cannot be added to DMV or CEC fleet snapshots. California Auto Outlook is published by Auto Outlook, Inc. on behalf of the California New Car Dealers Association. Data sourced from Experian Automotive. Access to the public report does not establish a license to the underlying Experian database.
How many new cars sold are zero-emission?
County pages, ZIP pages and the new-registration view show new light-duty sales by quarter from the California Energy Commission, with data through September 30, 2026. County and statewide figures give zero-emission sales against all new light-duty sales. ZIP figures count zero-emission sales only, because the agency publishes no ZIP total. The series start with the first quarter of 2025.
More detail
DMV records do not mark a registration as a new sale, so the agency infers one from a low odometer reading, a model year near the registration year and no earlier registration. It changed this method in 2023 and again in 2025, so earlier quarters may not compare with these. All quarters of the current year are preliminary and can change with each release until the agency finalizes them with the fourth quarter.
The agency assigns each sale by mailing address. Statewide figures leave out its "Out Of State" row, for cars registered in California to an address in another state, and ZIP figures leave out ZIP 99999, its code for an invalid ZIP. These sales counts are a separate source from the dealer association's registrations and the DMV snapshot, and are never added to them.
California Energy Commission (2026). California Energy Commission Zero Emission Vehicle and Infrastructure Statistics. Data last updated October 13, 2026. Retrieved November 8, 2026 from http://www.energy.ca.gov/zevstats.
Public places to meet the cars
The dealer directory contains seven California sales-dealer entries from Aston Martin's public directory and dealer pages, observed November 3, 2026. Addresses and coordinates describe business locations. Sales, pre-owned, service and parts tabs are not counted as separate dealers. This is one manufacturer's directory snapshot, not a complete dealer-license census or a vehicle-location dataset.
More detail
Public OpenStreetMap queries did not return usable records during collection, so no OSM dealer counts are displayed. No open bulk redistribution license was established for the manufacturer directory; only selected public business facts and source links are shown. All implemented enrichments were acquired without a data purchase or paid API.
Download the data
One CSV row for each of the 1,800 mapped ZIPs, built from the DMV snapshot of December 31, 2024. It has 18 columns: the ZIP, its city and first-listed county, vehicles by fuel, model years before 2011, model years 2024 and 2025, vehicles whose make the DMV does not name, and the largest named make with its count as a lower bound.
The file holds DMV counts only. You may reuse it at no charge with credit to Coast of Cars and a link to this page. The statistics page lists the key statewide numbers with their sources.
Download the ZIP table (CSV)Coverage in numbers
Full fuel, disclosed-make, model-year and duty-class breakdowns are available for all 1,800 mapped ZIPs. Statewide filters cover 116 cleaned make groups. Each make has a profile with ZIP rankings and fuel/model-year distributions.
| All retained DMV source counts | 31,141,961 |
|---|---|
| Mapped California ZIP counts | 30,118,651 |
| ZIP labels without a matched point | 1,003,092 counts |
| Points outside California | 20,218 counts |
| Pooled/unknown in mapped fleet | 3,470,400 (11.5%) |
More detail
The retained file has 678,982 aggregate rows and 129 source-wide make labels. Its embedded dates all say December 31, 2024, despite a newer portal resource label. The mapped subset has 630,001 source rows and 124 named make labels, grouped into 116 cleaned labels. We use the embedded snapshot date and keep excluded geography separate.
No individual plates, VINs, owners, leases, loans or payments are included. The data does not establish a complete supercar census, current road traffic, household ownership or wealth.