The youths were all born between April 1972 and March 1973 and had been studied longitudinally since birth in the Dunedin Multidisciplinary Health and Development Study. Each youth was provided a deck of 58 randomly sorted index cards, on each of which was printed a delinquent act. Confidentiality was explained, and the youths were asked to sort the cards according to whether their friends or other people their age that they knew had engaged in the act "never," "once or twice," or "three or more times." After completing this task, the participants were given a second identical set of cards and asked to report for themselves. In followup interviews, the researcher sought detailed descriptions of the circumstances of each act, related contacts with the police, and whether the subject was honest in reporting. The reliability and validity of the test instrument were analyzed. The test-retest reliability, internal consistency, known groups criterion validity, and convergent validity with reports of antisocial behavior from parents and teachers were all found to be adequate for recommending the instrument for use in social science research. In addition, findings in the predicted directions were obtained for the relationship between early delinquency and gender, family instability, history of behavior problems, social class, and intelligence quotient.