Poster presented at the 111 th Annual Convention of the American Psychological Association

and Ron Richardson for their assistance. This study examined the relationship between adolescent (10-17 years) criminal offending and adult (18-33 years) offending. The sample comprised 378 Canadian male offenders whose criminal trajectory was tracked for an average of 12.1 years (range = 5.8- 22.8), from adolescence into adulthood. Their mean age at the time of the most recent follow-up (March 17, 2001) was 27.5 years (range = 22.2- 33.5). The sample was divided into low and high rate offenders, based on adolescent convictions. Using rates of adolescent offending, separate Cox Proportional-Intensity Regression models were conducted within each group to predict adult offence rates for five types of offences (property, violent, sex, drug, and technical violations). The results indicated a high degree of continuity for property, drug, and sex offences. As well, late adolescence drug offences increased the likelihood of adult property offences. Both adult violent and technical offences were predicted by a range of adolescent offence types. The implications of these findings for advancing the criminal career concept are discussed