Drivers involved in 180 fatal collisions with Baltimore pedestrians were studied. Eighty-three drivers (46%) were judged to have been probably negligent and 66 (37%) were probably not negligent; negligence was not known for 31 (17%). Driver negligence was correlated with poor driving records. The study drivers had more points for traffic convictions than the average Maryland driver. Their driving records resembled those of drivers killed in crashes. Subsequently, 46 drivers (25%) were convicted of traffic violations. For the 22 drivers whose licenses were revoked, the median length of time until revocation was 8 months. Recommendations include chemical tests for alcohol of all drivers who kill pedestrians and swifter suspension of licenses of those drivers believed to pose a substantial hazard to society. Since the behavior of high-risk drivers may prove to be as difficult to modify as that of high-risk pedestrians, ultimate solutions probably lie in modifying roads, vehicles, and traffic patterns in order to reduce pedestrian injuries and deaths.