Cities nationwide have adopted so-called road diets to improve traffic safety, though they are sometimes met with intense opposition from motorists who fear that road diets will increase traffic delays. Road diets typically convert four-lane roadways with no left-turn lanes into streets with a center left-turn lane, two through-traffic lanes, and (often) bicycle lanes and right turn pockets at intersections. The resulting safety improvements are often dramatic. The Federal Highway Administration currently recommends that road diets should be applied to roadways with fewer than 20,000 average daily trips, but that cities should carefully consider whether to apply road diets above 20,000 average daily traffic (ADT). However, study of higher-traffic-volume road diets to inform decisions about them has been limited. In particular, there is scant evidence that safety benefits erode and traffic delays increase meaningfully above this threshold, though this is implied by the 20,000-ADT threshold. To address this literature gap, we examined the safety and traffic outcomes of high-traffic-volume road diets in Los Angeles, CA. To do this, we compared collisions on five high-traffic-volume road diet corridors with 16 similar multilane, untreated street segments. We found that collisions, injuries, and deaths were lower by 31.2% to 100%, depending on the measure, whereas traffic speeds were lower by about 6.7% (peak) to 7.9% (off-peak). We concluded that in Los Angeles higher-traffic-volume road diets appeared to significantly increase safety with only minor effects on traffic speeds.