The age distribution of crime is one of the most replicated facts of criminology with many studies indicating that offending peaks during adolescence and gradually declines during adulthood. This paper tests several hypotheses regarding the age-crime relationship derived from three theoretical perspectives: (1) age distribution of crime does not vary across place and time; (2) changes in life circumstances directly influence criminal behavior; and (3) aggregate age-crime distribution is actually a combination of several distinct groups of offenders, each following different developmental trajectories. The study examined the extent to which life circumstances affected crime after taking enduring individual differences and the extent to which effects of life circumstances varied between offenders into consideration. The study also examined the extent to which variation in crime by age at the aggregate level was due to age-graded differences in life circumstances and the distribution of offender types in the population. The study utilized two datasets: official data and self-report individual-level data covering the 60 years from age 12 to age 72. Findings from the analysis indicated (1) a strong relationship between age and crime at the individual level that varied across individuals, (2) even after between-individual differences in criminal propensity were accounted for, life circumstances still affected offending, and (3) strong support was shown for the idea that the association between life circumstances and crime resulted not solely from social selection, but also from social causation. Tables, figures, and references