Melatonin is a pleiotropic molecule with many cellular and systemic actions, including chronobiotic effects. Beneficial effects are widely documented concerning the treatment of neoplastic diseases in vivo as well as reductions in viability of cultured cells from melanoma, one of the most aggressive cancers in humans. However, studies of its effects on non-tumor cells in vitro have not focused on viability, except for experiments aiming to protect against oxidotoxicity or other toxicological insults. Furthermore, there is no agreement on the range of effective melatonin concentrations in vitro, and the mechanisms that reduce cell viability have remained unclear. Tumor cell-specific increases in the production of reactive oxygen and nitrogen species (ROS/RNS) may provide a possible explanation. Our aim was to analyze the potential inhibition of tumor (B16 melanoma 4A5) and non-tumor cell (3T3 Swiss albino) viability using a wide range of melatonin concentrations (10(-11)-10(-2) M), and to determine whether intracellular ROS enhancement was involved in this process. In the absence of fetal bovine serum (FBS), low melatonin concentrations (10(-9)-10(-5) M) reduced the proliferation of melanoma cells with no effect in fibroblasts, whereas, in the presence of FBS, they had no effect or even increased the proliferation of both fibroblast and melanoma cells. Melatonin concentrations in the upper millimolar range increased ROS levels and reduced the viability of both cell types, but more markedly so in non-tumor cells. Thus, low melatonin concentrations reduce proliferation in this specific melanoma cell line, whereas high concentrations affect the viability of both tumor (B16 4A5 melanoma) and non-tumor (3T3 fibroblasts) cells. Increased ROS levels in both lines indicate a role for ROS production in the reduction of cell viability at high-but not low-melatonin concentrations, although the mechanism of action still remains to be elucidated.