The age at which people begin to offend and the variety of different offenses that they commit while active have independently received a considerable amount of attention from researchers. However, there has been little attempt to connect these two relatively distinct lines of inquiry. Recently, two prominent criminological theories, Gottfredson and Hirschi's general theory of crime and Moffitt's taxonomy of offending behavior, have advanced formal theoretical models that include predictions about the relationship between onset age and offense versatility. The authors test the validity of these predictions using data from the second Philadelphia birth cohort study conducted by Tracy, Wolfgang, and Figlio. In support of the predictions from these theories, the authors find a relationship between onset age and offense versatility. More detailed analyses, however, show that this association vanishes after controlling for age. Thus, there was a tendency for offenders to become more specialized in their offending over time regardless of the age at which they initiated offending. The theoretical implications of this finding are discussed.