The study found autonomous vehicles had a 68% lower crash rate than human drivers.
"What we're seeing is a nice pattern - fewer crashes overall, less egregious types of crashes, and less severe crashes. Our findings line up pretty well with the published research from Waymo," said Eric Teoh, one of the researchers behind the report.
The study examined Waymo crash data in four cities between 2021 and 2024. According to the findings, crash rates were 76% lower than human drivers in Phoenix and 71% lower in Los Angeles. In San Francisco, the difference was smaller, with autonomous vehicles experiencing a crash rate 35% lower than human drivers.
Austin was the only city included in the study where Waymo vehicles recorded a higher crash rate than human drivers, at 4% more crashes.
MORE: SF may soon fine autonomous vehicle companies for impeding emergency response
Teoh said there was not enough data from Austin to provide a fully accurate comparison.
"We observed that in, say, front to rear crashes, the Waymo is almost never the striking vehicle. And even though it gets rear-ended more than it strikes other people, it also importantly gets rear-ended less than human drivers do," Teoh said.
In a statement, Waymo welcomed the findings.
"This study is an important contribution to a growing body of research on AV safety data, and we're looking forward to additional analyses from the scientific community," the company said. "More high-quality data is fundamental for building public trust in this technology and ultimately making roads safer for everyone."
Not everyone agrees the study settles the question of autonomous vehicle safety.
MORE: July 4 Waymo gridlock in SF after dozens choke streets, some lose power, 1 ran over lit fireworks
"I don't think people are going to buy it," said Ahmed Banafa, a professor at San Jose State University.
Banafa said the study's scope was too narrow because it focused on crashes and did not account for incidents in which autonomous vehicles may have acted unsafely without causing a collision.
He cited issues such as San Francisco's Fourth of July traffic logjam, which he said would not be included in the study because autonomous vehicle companies are not required to report such incidents.
"Is that part of the reporting or just the driving that you didn't hit a car, you didn't stop, you didn't scratch though? So they are very selective, let me put it that way, in what they are reporting. And that's not helping. They have to be open about everything, the good and the bad and the ugly," Banafa said.
As autonomous vehicles become more widely available, both Banafa and the IIHS said changes to how safety data is reported and shared could help provide a more complete picture of the technology's performance on public roads.