The sample of 378 subjects all had contact as juveniles with the Toronto police from July 1 to December 31, 1970. The information on each subject dates from original contact with police; if any information as a result of police contact was available prior to the 1980 entry, it was included in the file along with all other information up to December 1974. Subjects' files were updated to include information on adult offenses. Offenses were categorized as status offenses, nonviolent offenses, violence to property, and interpersonal violence. Since data on the exact number of youths involved in offenses were not available, group activity was defined as the involvement of two or more juveniles in an offense. The distribution of offenses committed by the subjects indicates a fairly high degree of group involvement (62.9 percent). The hypothesis that group involvement in delinquent acts increases the probability of police arrest appears to have greater validity for criminal code offenses, particularly violent property offenses. An examination of the pattern and potential implications of group violence over time indicates that juveniles are not prone to greater involvement in group violence as their careers progress; moreover, examination of the means and standard deviations for career violence shows that youths are slightly less prone to group violence in the second half of their careers, while demonstrating a somewhat greater proclivity to commit violent acts alone. Implications of the findings and suggestions for future research are offered. Tables and bibliographic entries are provided.