The biochemical effects on catecholaminergic systems of the dopamine antagonist amisulpride and the dopamine agonist bromocriptine were evaluated in a double-blind, randomized, crossover study in children with autistic disorder. Plasma levels of dopamine, norepinephrine, and epinephrine; urinary concentrations of homovanillic acid (HVA), vanillyImandelic acid (VMA) and 3-methoxy-4-hydroxyphenylethylene glycol (MHPG); and plasma prolactin were measured. At doses of amisulpride and bromocriptine that had the expected opposing effects on plasma prolactin, the drugs' effects on the catecholaminergic systems were similar. Both agents unexpectedly lowered urinary HVA (total, free, and sulfated) although only amisulpride decreased the HVA levels significantly. This paradoxical decrease in HVA levels suggests that both dopamine agonists and antagonists could act on autoreceptors or presynaptic dopaminergic receptors in the central nervous system. There was no significant action of either drug on plasma epinephrine, urinary VMA, or urinary MHPG, suggesting that neither drug altered norepinephrine or epinephrine systems; however, a weak increase of plasma norepinephrine occurred after amisulpride treatment, consistent with effects observed with other neuroleptics. Neither agent altered plasma dopamine, suggesting that peripheral dopamine metabolism was unchanged. The clinical effects of amisulpride and bromocriptine have been reported to be unexpectedly complementary. The complementary clinical effects of a dopamine agonist and dopamine antagonist might speculatively be related to their similar action on dopamine autoreceptors, leading to a correction of the dopaminergic hyperactivity that has been postulated in autistic children.