Gender differences in criminal offending patterns have been the subject of much research, although most of this work focuses on pathways to juvenile male offending. Silverthorn and Frick (1999) recently put forth a gender-specific model of pathways to juvenile offending for girls in which they claim that girls’ antisocial behavior does not manifest until puberty, leading to a delay in offending onset for girls even if the girls have the same high-risk backgrounds as boys who display early-onset offending behavior. The current article presents a preliminary empirical test of this model using longitudinal data collected from 220 African-American individuals from birth to their mid-30's. Data were drawn from the Philadelphia portion of the National Collaborative Perinatal Project, which collected information about biological, psychological, and sociological variables. Additionally, a follow-up was conducted in 1998 in which criminal history variables were examined. Results of statistical analyses indicated that boys and girls were equally likely to engage in early-onset offending patterns. However, female late-onset offenders were similar to male early-onset offenders on 9 of 10 risk factors. Moreover, while male early-onset offenders had more severe criminal offending when compared with female and male late-onset offenders, the male early-onset offenders did not differ significantly from the early-onset female offenders. The findings suggest that Silverthorn and Frick (1999) may have underestimated the prevalence of early-onset-life-course persistent offending in female juveniles. Continued research is warranted on the small subgroup of early-onset female juvenile offenders. Tables, references, notes