Twenty-eight children took part in a double-blind, placebo-controlled, crossover study of fenfluramine and methylphenidate. Fenfluramine dosage was gradually increased to a standardized dose of 1.5 mg/kg per day, whereas methylphenidate was given in doses of 0.4 mg/kg per day. The children were assessed on laboratory tests of selective and sustained attention, visual matching, and color matching, during which seat activity was monitored automatically. Results showed fenfluramine to be superior to placebo on the memory task, whereas methylphenidate reduced commission errors on a continuous performance test. Methylphenidate caused shorter response times, and fenfluramine caused increases, on two of the tests. Examiner behavior ratings indicated significant improvements with both drugs on the domains of attention, activity level, and mood. These findings, together with those from a companion clinical study, suggest that the drugs may have contrasting mechanisms of action, but both appear to have useful clinical effects in these children.