Birth cohort studies provide the best, if not the only, basis upon which to make significant probability statements about delinquent and criminal events. They also permit detailed and precise description of transition probabilities, that is, moving from one type of offense to another. The original Philadelphia study involved analysis of a cohort of males born in 1945 who lived in Philadelphia at least from their 10th to their 18th birthdays. Data were gathered on 9,945 boys. Since 1968, a 10-percent random sample of the original cohort was followed until the males reached age 26 or 30. The followup showed the probabilities of being arrested, of being a serious assaultive offender, of being a chronic offender, and of committing any Index offense by age-specific rates. The article discusses the probabilities of desistance at each offense number and the offense transitions by offense types and by number of offenses. Changes from juvenile to adult status are captured, and the importance of gang membership as a contributor to serious juvenile and adult criminality are documented in bivariate and covariate analyses. Finally, in a log-linear model the significant relationship between being a victim of crime, particularly as a juvenile, and becoming a serious violent offender as an adult, is noted. A total of 27 footnotes and 6 appended tables are provided. (Author summary modified)