Each of 28 nonautistic children with attention-deficit hyperactivity disorder and mental retardation received placebo, methylphenidate (0.4 mg/kg/day), and fenfluramine (gradually increased to 1.5 mg/kg/day) for 4 weeks each in a double-blind, crossover design. Teacher ratings indicated significant improvements with both active drugs on subscales designated as Conduct Problem, Hyperactivity, and Irritability, but methylphenidate alone produced improvements on an Inattention subscale. Parent ratings indicated significant improvements with both drugs on subscales labeled Hyperactivity, Motor Excess, and Conduct Problem. Fenfluramine alone caused improved parent ratings on Irritability and Inappropriate Speech, and on Conners' Abbreviated Symptom Questionnaire. Unlike a previous study, subgroup analyses failed to show a significantly better clinical response to methylphenidate for subjects with higher mental ages, although children with higher IQs responded better than those with IQs less than 45. The active drugs had contrasting effects on heart rate and blood pressure. Fenfluramine caused significant weight reductions relative to both placebo and methylphenidate. These findings suggest that both methylphenidate and fenfluramine have useful, but somewhat different, clinical effects in certain children with attention-deficit hyperactivity disorder and mental retardation.