The potential cytotoxicity of the melanogenic intermediates DOPA, (L-3,4-dihydroxyphenylalanine) and DHI (5,6-dihydroxyindole) has long been recognized and exploited as a targeting concept in experimental melanoma therapy. In recent years, however, a novel branchpoint in the melanin biosynthetic pathway has been shown to divert the metabolism of DOPAchrome to a carboxylated derivative termed DHICA (DHI-2-carboxylic acid) rather than to DHI. In order to evaluate the biological implications of this regulatory control, we have reexamined the inherent cytotoxicity of DHICA versus DHI on different cell lines. We found that under the usual conditions of the biological assay, the apparent cytotoxicity of the two indoles reflect their instability in the culture medium, the less stable DHI being generally more toxic than DHICA to melanoma cells and nonmelanocytic cells. Moreover, the observed cytotoxic effects increased with the time of incubation and were markedly reduced by the addition of catalase to the medium, suggesting that they were probably due to the generation of reactive oxygen species (particularly H2O2) during the autoxidation of the melanin precursors outside the cells. To circumvent this problem, we then tested the diacetylated derivatives of DHI and DHICA (DAI and DAICA) which are sufficiently stable until taken up into the cells whereupon they may be converted by endogenous esterases back to the parent indoles. Although DAI proved to be cytotoxic for nonmelanocytic cells, it had no detectable activity on melanoma cells, whereas DAICA showed no effect on any of the cells examined. These results, when combined with other studies, point to a reconsideration of the inherent cytotoxicity of the 5,6-dihydroxyindoles, as well as DOPA, to melanin producing cells.