After explaining that clinical subtyping is a procedure for the classification of individuals who all share more than one common characteristic with diagnostic or treatment relevance, the authors discuss the typological data collected from 600 adolescents presenting for marijuana treatment. Participants came from the Cannabis Youth Treatment (CYT) study and were classified according to gender, substance use onset age, family history, externalizing disorders, internalizing disorders, and temperament. By comparing each of the classified subgroups according to substance use frequency, substance abuse problems, support for substance use, family and school problems, and negative peer associations, the authors found support for their argument that classification schemes have valid explanatory and predictive powers. The authors found that each of the categorical classification schemes was able to differentiate significant subtypes for either all or some of the construct validation measures. Furthermore, the authors found that externalizing disorders, onset age, difficult temperaments, and internalizing disorders among CYT program participants added unique variance to discrimination after the effects of the other two subtypes had been removed. While a follow-up study conducted 12-months after the program ended demonstrated that there were no differences among program participant subtypes concerning substance use frequency, the adolescents with high levels of externalizing and internalizing disorders did experience more frequent substance use problems. Tables, references