About the numbers
Every stat on the board links to its source. This page explains what each number means — and, just as important, what it doesn't.
Deaths caused
Fatalities in crashes that investigators, police reports, or federal filings attribute to the robotaxi's own driving while it was operating with no human driver. As of the last update, no death has been attributed to either Waymo's or Tesla Robotaxi's driving.
Involved in fatal crashes
Fatal crashes a driverless robotaxi was part of at all — including crashes caused entirely by human drivers. Waymo's two: a January 2025 San Francisco pileup in which a speeding driver struck stopped traffic that included an empty Waymo, and a September 2025 Tempe crash in which a motorcyclist rear-ended a slowing Waymo and was then hit by a human hit-and-run driver. Neither was Waymo's fault, but honesty requires showing the involvement count rather than pretending the crashes didn't happen.
Injury crashes & all reported crashes
Since 2021, federal law (NHTSA's Standing General Order) requires automated-driving companies to report qualifying crashes. "Injury crashes" counts reports that include any injury, minor to serious. "All reported crashes" is every filed report, most of which are minor — fender-benders, sideswipes, and being rear-ended by human drivers.
Two big caveats. First, NHTSA records involvement, not fault — a robotaxi rear-ended at a red light still generates a report. Second, exposure differs enormously: Waymo has driven over 220 million driverless miles with thousands of cars since 2020, while Tesla's Robotaxi service launched in June 2025 with a few dozen vehicles and does not disclose mileage. Raw counts are not safety rates. Waymo's own published analysis (through March 2026) reports 82% fewer injury crashes per mile than comparable human drivers; Tesla's per-mile rate cannot be independently computed because its mileage is undisclosed.
Cities in service
Metro areas where the public can hail a fully driverless ride today — no human safety driver in the car. Announced-but-not-launched cities are noted separately. Tesla's SF Bay Area service is excluded because California permits currently require a human safety driver there.
Cars on the road
Best-available count of driverless vehicles actually deployed. Waymo last confirmed ~2,500 in the US (late 2025); independent trackers estimate 3,500+ as of mid-2026. Tesla does not publish fleet numbers; the figure shown comes from vehicle registrations and independent tracker counts.
What about consumer Teslas on Autopilot or FSD?
Not counted. Privately owned Teslas running Autopilot or "Full Self-Driving (Supervised)" have a licensed human driver responsible at the wheel — a different product with its own crash record. This site tracks driverless taxi services only. The two are frequently conflated in news coverage; the scoreboard keeps them separate.
How updates work
The board refreshes weekly from NHTSA crash-report data, company announcements, and press reporting, with a source linked beside every number. Numbers only change when a source supports the change. Spot an error? Email us with a link and we'll correct it.
For media & researchers
You're welcome to cite, quote, or republish the numbers on this site with attribution to RoboTaxi Report and a link to robotaxireport.com. Suggested citation: "RoboTaxi Report, the Waymo vs Tesla robotaxi scoreboard (robotaxireport.com), data as of [date shown on the board]."
The full dataset behind the board is open: download the current data as JSON. Every figure carries its original source (NHTSA, IIHS, company disclosures, press reporting) — for primary-source citations, use the source links beside each number on the board. Questions, interview requests, or corrections: hello@robotaxireport.com.
Independence
RoboTaxi Report is independent and not affiliated with, sponsored by, or endorsed by Waymo, Tesla, or any autonomous-vehicle company.