The study samples were drawn from diverse settings in Canada (seven studies), the United States (two studies), and the United Kingdom (1 study). The analyses were conducted on a dataset that combined the 10 study samples. Age was measured at the time of release from an institution, except for one of the samples, for which age was measured at the time of sentencing. Sexual recidivism was measured according to the definitions used in the original studies, which in most cases was either charges (four studies) or convictions (5 studies). One study included readmissions to psychiatric facilities as recidivism. Preliminary analyses (logistic regression) measured the association between age and sexual recidivism. Based on the offenders' predominant victim choice, the sample was divided into those who sexually victimized adult women (rapists, n=1,133); child molesters who victimized unrelated children (extrafamilial child molesters, n=1,411); and child molesters who victimized only related children (incest offenders, n=1,207). Excluded from the classification were 47 offenders who victimized both adult women and unrelated children and 875 offenders for whom victim information was not available. The study found that rapists were younger than child molesters, and the recidivism risk for rapists steadily decreased with age. In contrast, extrafamilial child molesters showed relatively little reduction in recidivism risk until after the age of 50. The recidivism rate of intrafamilial child molesters was generally low, except for the intrafamilial offenders in the 18-24 age group, whose recidivism risk was comparable to that of rapists and extrafamilial child molesters. The discussion focuses on developmental changes in sexual drive, self-control, and opportunities to offend. 2 tables, 2 figures, and 47 references