Trends in fatal crashes and injuries per 100 crashes were evaluated in Saving Lives cities and the rest of Massachusetts between March 1984 and February 1993. The evaluation involved the following components: fatal and injury crash monitoring, direct observation of safety belt use and speeding, telephone surveys, and traffic citation monitoring. In Saving Lives cities, fatal crashes declined from 178 during the 5 preprogram years to 120 during the 5 program years. Fatal crashes involving alcohol in program cities dropped from 69 to 36, a 42-percent decline. The number of fatal crashes that involved speeding drivers decreased from 68 to 33 in program cities, and the number of pedestrian fatalities declined from 45 to 33. Safety belt use increased in program cities from 22 percent in 1989 to 29 percent in subsequent years. By the end of the fifth program year, 54 percent of respondents between 16 and 19 years of age and 40 percent of adults in program cities were aware of the Saving Lives Program. The proportion of adults in program cities who believed police would stop drunk drivers and speeders did not significantly increase, while the proportion of teenagers who believed the license of a person caught drinking and driving could be suspended increased from 61 percent to 76 percent. During the 5 program years relative to the previous 5 years, program cities experienced a 33-percent decline in fatal crashes, from 178 to 120. In comparison cities, the number of such crashes increased from 120 to 121. Overall, during the Saving Lives Program, Massachusetts had the lowest traffic fatality rate per vehicle miles driven of any State in the United States. The authors conclude that interventions organized by multiple city departments and private citizens can reduce drinking and driving, related driving risks, and traffic deaths and injuries. 49 references, 1 table, and 4 figures