Focuses on limitations in the ways in which child and adolescent psychotherapy has been studied. Limitations pertain to narrow sampling of the types and characteristics of clinical dysfunction and treatments and many sources of variation (child, parent, and family characteristics) that are likely to influence outcome. I discuss the relation between limited conceptualization of treatment and the methods of study and the resulting knowledge about treatment. Recommendations are made in relation to assessment, parametric studies of treatment, models of treatment delivery, and the study of mechanisms of therapeutic change as the means to develop effective treatments, identify children for whom they are effective, and understand how treatments operate.