This population-based case-control study contrasted nutrient intakes of 41 women with cutaneous malignant melanoma to those of 297 women sampled from the same community (Brisbane, Australia). Diet was assessed by a comprehensive food frequency questionnaire. The strong inverse relation we observed between high intakes of polyunsaturated fatty acids and melanoma (P < 0.01) adds sufficient weight to prior findings for this persisting causal hypothesis to be abandoned. A relatively strong association with alcohol was observed: Women drinking 20 g or more (two or more drinks) daily had 2.5 (odds ratio) times the risk of melanoma as nondrinkers (95% confidence interval, 0.87 to 7.4). However, previous data are inconsistent. With regard to potential protective factors, our data fit with prior speculation that antioxidants (beta-carotene and vitamin E), zinc, and iron warrant further investigation. The nonsignificant (P = 0.18) 40% reduction in risk seen for those eating the most fish (> or = 15 g daily versus < 5 g) suggests the effects of marine oils and omega-3 fatty acids may deserve specific attention.