The present experiments were designed to test the hypothesis that the long-term effects of p-chloroamphetamine (PCA) on serotonergic neurons in rats are mediated by a neurotoxic metabolite. The effects of well-known inducers and an inhibitor of hepatic microsomal drug-metabolizing enzymes on the PCA-induced decreases in brain levels of 5-hydroxytryptamine (5-HT) and tryptophan hydroxylase activity and the half-life of PCA in brain were examined. All of these modified the half-life of PCA in a predictable manner: 3-methylcholanthrene and, to a much lesser extent, phenobarbital decreased the half-life of PCA while piperonyl butoxide markedly increased it. Fluoxetine, an inhibitor of 5-HT uptake, also increased the half-life of PCA in brain. In addition, fluoxetine blocked the long-term effects of PCA on 5-HT levels and tryptophan hydroxylase activity. Of the classical metabolic tools, only 3-methylcholanthrene provided complete protection from the long-term, neurotoxic effects of PCA. Although the mechanism of this protection is unknown, it is not mediated by a blockade of 5-HT uptake since 3-methylcholanthrene did not decrease the synaptosomal uptake of 5-HT. Piperonyl butoxide pretreatment markedly increased the half-life of PCA in brain, but it failed to modify consistently the effects of PCA. The results indicate that the long-term, neurotoxic effects of PCA are not mediated by the major, hepatic metabolites of the drug, but leave open the possibility of a minor, neurotoxic metabolite.