Analysis of career desistance was undertaken for the 6,287 cohort members who experienced official police contact. Results indicate that a smaller proportion of adult offenders among delinquency dropouts than among youth with more stable delinquent careers. This finding held regardless of race, socioeconomic status, gender, or drop-out period. Logistic regression also reveals that cessation during youth, measured by age at final police contact, had a significant effect on adult offending. Results could have significant implications for the development of effective criminal justice policies. The differential prevalence rates of adult crime suggest that contrasting strategies for law enforcement, judicial decisions, and correctional treatment may be warranted. For example, official intervention in career progression may assist in deterring potentially chronic offenders from persistent criminal careers. While a stepwise logistic regression model was able to predict adult offending with 72 percent accuracy, the rate of false positives was sufficiently high to present ethical dilemmas in its application to selective incapacitation decisions. 6 tables, 5 notes, and 60 references.