In the past 5 years, we have witnessed the continuation of important trends in clinical research that began earlier in the decade. With regard to the treatment of specific disorders in children and adolescents, the most significant developments have been the examination of the tricyclics for the treatment of depression and the initiation of controlled studies for the treatment of Tourette syndrome. Unfortunately, the findings from the depression studies have been uniformly negative, and the results of research on both depression and tic disorders show a relatively high rate of placebo responsivity, which raises nagging questions about the role of case reports and open trials. Another important trend in pediatric psychopharmacotherapy is the search for substitutes for the neuroleptics. Potential candidates include agents such as lithium, naltrexone, fenfluramine, clonidine, and carbamazepine. The most underresearched disorders are a combination of the least common (e.g. schizophrenia, mania) and those that are apparently perceived as less serious (e.g. sleep disorders, certain anxiety disorders). Not surprisingly, the most studied disorder and treatment is hyperactivity and stimulant medication, respectively. Considerable progress has been made in understanding the social implications of the associated symptoms and their response to stimulant drugs, aided greatly by the use of direct observation procedures. Researchers are beginning to attend to the implications of comorbidity for assessing response to medication. There has been additional confirmation of efficacy of stimulant treatment for preschoolers and adolescents. Dose-response issues remain to some extent unresolved, the primary impediments being interpretive misconceptions associated with trend analysis, an overreliance on the syndromal perspective and too little attention to target behaviors and their clinical implications, and the failure to operationalize the minimal effective dose with regard to the normalization and supranormalization of target and collateral behaviors. Disagreement over whether hyperactivity is a learning or a behavior disorder (or both) and what academic underproductivity means clinically and socially is also impeding progress. With regard to developmental disorders, controlled studies indicate that fenfluramine and naltrexone are effective for managing associated symptoms in some individuals. However, given the limited amount of research on these agents, their status as clinically useful palliatives must be considered tentative.(ABSTRACT TRUNCATED AT 400 WORDS)