We did not nd clear evidence for two separate life-course persistent and adolescent-limited groups of delinquents (Mott, 1993). In contrast to Laub and Sampson (2001), we found that early factors predicted desistance, and we did not nd that life events such as having a wife or a girlfriend or being in the military predicted desistance from oending. Our results are concordant with bonding theories like those proposed by Hawkins’s social development theory (Hawkins et al., 2003), which also emphasizes promotive factors. Our results are also concordant with ornberry and Krohn’s (2003) focus on bonding, and the importance of parenting, school problems, and peer delinquency.