The authors report the results of latent class analyses and latent class transition analyses of antisocial behavior risk factors among 137 youths participating in a juvenile diversion program. The study examined the youths' latent classifications using baseline and 1-year follow-up measures of family, peer, education, and mental health risk factors. Latent class transition analyses were conducted to determine the stability and change of latent class membership across two time points. For both baseline and follow-up measures, latent class analyses suggested that two latent classes were most appropriate for characterizing the youths using the four domains of risk factors. One class reflected youths reporting few problems in any risk area; the second class reflected youths reporting problems in many of the risk domains. The latent class transition analyses indicated that about two thirds of the youths experienced similar problems over time. The results emphasize the heterogeneity of this study's sample. Policy implications and intervention service needs are discussed. (Contains 2 figures and 3 tables.)