1-(3,4-Methylenedioxyphenyl)-2-aminopropane (MDA) produces an effect in humans that is somewhat similar to both a hallucinogen and a central stimulant. We have previously shown that R(-)-MDA but not S(+)-MDA produces a stimulus effect in animals similar to that of the hallucinogen 1-(2,5-dimethoxy-4-methylphenyl)-2-aminopropane (DOM), whereas S(+)-MDA but not R(-)-MDA produces a stimulus effect similar to that of the stimulant amphetamine. Others have suggested that the stimulus effects of the two MDA isomers may be much more similar than our results would indicate. To help clarify this issue, we have shown that rats (n = 8) can be trained to discriminate 1.25 mg/kg of R(-)-MDA (ED50 = 0.99 mg/kg) from 1.25 mg/kg of S(+)-MDA (ED50 = 0.80 mg/kg) versus 0.9% saline with use of a three-lever operant procedure. To accomplish this task it was necessary to institute a separation of 4 days between injections of the isomers. In tests of stimulus substitution, the administration of various doses of DOM (ED50 = 0.47 mg/kg), S(+)-DOM (ED50 = 2.23 mg/kg) and mescaline (ED50 = 16.04 mg/kg) produced dose-related responding on the R(-)-MDA-appropriate lever. In contrast, the injection of various doses of S(+)-amphetamine (ED50 = 0.46 mg/kg) and cocaine (ED50 = 6.40 mg/kg) produced dose-related responding on the S(+)-MDA-appropriate lever. The administration of racemic MDA, at twice the isomer training dose, resulted in the animals dividing their responses closely between the R(-)-MDA- and S(+)-MDA-designated levers. In antagonism tests, the serotonin-2 antagonist pirenperone blocked the stimulus effect of 1.25 mg/kg of R(-)-MDA but had no effect on the stimulus effect of 1.25 mg/kg of S(+)-MDA. Taken together, these data support the contention that each optical isomer of MDA can produce a markedly different stimulus effect.