Immunohistochemical methods were used to analyse the distribution and morphology of serotonergic axons in normal macaque monkeys and in monkeys given (+/-)3,4-methylenedioxymethamphetamine. In untreated monkeys, we observed two morphologic classes of serotonergic axon terminals, which differ in regional and laminar distribution. These two axon types, fine and beaded, correspond to the serotonergic axon types which have been described in the rat. In 3,4-methylenedioxymethamphetamine-treated monkeys, there is a profound loss of serotonergic axon terminals, yet some are consistently spared. The surviving axon terminals are nearly all of the beaded type; in contrast, fine serotonergic axons are markedly reduced in density. There are regional differences in the magnitude of denervation, which reflect differences in the distribution of these two types of serotonergic axons in controls. The present study demonstrates that 3,4-methylenedioxymethamphetamine has differential neurotoxic effects on fine and beaded serotonergic axons. These results indicate that in the primate there are two distinct classes of serotonergic axon terminals, which differ in morphology, distribution, and vulnerability to psychotropic drugs. We hypothesize that in the primate, as demonstrated in the rat, these two classes of serotonergic axon terminals may arise from different raphe nuclei. In both rodent and primate, the dorsal and median raphe nuclei give rise to parallel ascending serotonergic projections, which are likely to have different pharmacologic properties and functions.