Recent studies have suggested the inhibition of cyclooxygenase-2 (COX-2) as strategy to prevent colorectal cancer. In this study, the cytostatic and cytotoxic effects of different non-steroidal anti-inflammatory drugs (NSAIDs), all of them are reported COX inhibitors, were investigated in human skin melanoma A-375 cells. Using BrdU-cell proliferation assay, we showed that 50 and 100 microM of celecoxib (CEL) reduced proliferation of the melanoma cells at 72-h incubations by 34.0% and 82.7%, respectively. As determined by Toxilight-cytotoxicity assay, the drug was only toxic to the cancer cells at 100 microM. Indomethacin (IND) also inhibited the cell proliferation by about 40% at 240 and 480 pM and was only slightly toxic to the melanoma. Neither aspirin (ASP) nor piroxicam (PIR) exhibited cytostatic or cytotoxic effect on the cancer cells. Combinatory effects of the above NSAIDs with dietary docosahexaenoic acid (DHA) on inhibiting growth of the melanoma cells were further elucidated. Each of the NSAIDs, at doses 10-480 pM, was incubated simultaneously with the melanoma cells and 160 pM of DHA for 72 h. Results from MTT assay showed that both CEL and IND, starting from 20 microM. exhibited additive effects on the DHA-induced growth inhibition. ASP also enhanced the DHA-induced growth inhibition by 42.8% at 480 microM. To our surprise, although PRX did not suppress the melanoma growth, the drug at 40-240 microM enhanced the DHA-induced growth inhibition by 15.9-66.4%, respectively. Results from these studies suggest that the anticancer effects of NSAIDs may not be explained solely by their COX-inhibitory activities. Further studies are therefore required to understand their modes of action, before they could be used alone or in combinations with other agents for cancer chemoprevention.