Independent evidence for safer streets

Zero is the goal. Evidence is how we get there.

Evidence for Zero turns California public collision records into maps, local profiles, and independent research—helping communities see where people are being harmed, where the records fall short, and where safer systems are needed.

Latest 30 available data days: Jul 8, 2026 to Aug 6, 2026. Blue marks record-only crashes, yellow and orange mark reported injuries, and red marks fatal crashes. The map plots a performance-limited sample and reports full matching and missing-coordinate totals.

Explore the public record

Start with the question you have.

The project is organized around durable research paths, not a single dashboard. Move from a broad pattern to the place, road, map, or report that can support it.

01

Start with a place

Search a California city or ZIP area, then compare its reported crash patterns with statewide context.

Find a place
02

Follow a road

Review maintained highway corridor profiles with reporting windows, mapped records, city tables, and limitations.

Browse road profiles
03

Inspect the geography

Use the research map to filter reported crashes by date, place, severity, road, vehicle, and location coverage.

Explore the map
04

Read a finding

Open maintained reports with a defined question, fixed window, supporting evidence, methodology, and citation.

Browse reports

Latest available briefing

The Pulse follows the records, not the calendar.

The Pulse is a rolling 90-day view of the newest California crash records available in the warehouse. Its source date and reporting window are shown explicitly.

Read the latest Pulse
Source data through
Aug 6, 2026
Reported crashes
70,708
ZIP areas represented
1,668

Reporting window: May 9, 2026 to Aug 6, 2026. Counts may change when source agencies revise public records.

Support the public work

Help keep California crash data readable.

Sponsorship helps sustain careful data cleaning, public maps, maintained research pages, and transparent source notes. Sponsors are disclosed publicly and cannot influence the records, methods, findings, or corrections.

Source and independence

Useful evidence starts with visible limits.

The project uses California Highway Patrol California Crash Reporting System public records. It is independent civic research operated by Meru AI and is not affiliated with or endorsed by CHP or any government agency.

Crash counts are not exposure-adjusted risk rates. Source records, road labels, city labels, and derived locations may be incomplete or revised.

Read sources, methodology, and privacy

Use the work responsibly.

Journalists, educators, researchers, planners, and community organizations can cite maintained pages, build an embed, or contact the project about a local question.