All of the studies reviewed were conducted since 1970 and contain comparable pretreatment and posttreatment arrest data. Six were conducted in New York City; one in New Jersey; four in Boston, Detroit, San Antonio, and Philadephia; and one was a national study of treatment programs. The methodological problems of the studies are discussed, including exclusion of some subjects, the small sample for followup, absence of control for length of time in treatment, pretreatment and postreatment arrest period inadequacies, and inadequate arrest data. Findings of 7 of the 10 studies concerning methadone maintenance treatment showed a higher pretreatment arrest rate than a posttreatment arrest rate, while all the studies of residential drug-free treatment showed similar results. Issues raised by the 12 studies are discussed focusing on profiles of treatment program clients, profiles of treatment programs and impacts, the use of control or comparison groups, and prediction of success in treatment. To demonstrate the advantages of arrest data, particularly statewide official arrest data, for followup of all entrants into treatment programs, the paper discusses the disadvantages of other types of posttreatment outcome data, as well as the theoretical advantages and disadvantages of arrest data, particularly in view of findings from the New Jersey study. Suggestions for future research include obtaining arrest data from official sources, using a pretreatment arrest period of from 2 to 5 years, presenting of pretreatment and posttreatment data as the number of arrests per 100 person-years, dealing with length of time in treatment, selecting cohorts of entrants and including all those exposed to treatment, allowing for dropout and reentry data for methadone treatment programs, and considering the characteristics of treatment programs and clients. Tables and 30 references are included. For the full text of the drug use and crime report, see NCJ 40293.