Ecological and case-control studies have demonstrated a positive correlation between consumption of fat and the risk of prostate cancer. Two recent human studies have focused on alpha-linolenic acid as a risk factor for prostate cancer. Animal experiments have shown that dietary omega-6 polyunsaturated fatty acids have generally stimulated tumour development, whereas omega-3 polyunsaturated fatty acids have diminished it. The aim of our study was to investigate the association between these fatty acids and the subsequent risk of prostate cancer. Blood donors to the Janus serum data bank in Norway, who later developed prostate cancer, were matched to blood donors without prostate cancer (141 matched sets); the proportional level of fatty acids measured before diagnosis in the donors' serum was examined. The risk of later prostate cancer was analysed by conditional logistic regression. Increasing risk for prostate cancer was found with increasing quartiles of palmitoleic, palmitic and alpha-linolenic acid. An inverse risk association was found with increasing levels of tetracosanoic acid, for the ratios of linoleic to alpha-linolenic acid and arachidonic to eicosapentaenoic acid. There was no clear association between the risk effect of total omega-3 and total omega-6 fatty acids. There were no indications of a relationship between fatty acids and more aggressive cancers. Our results verify recent findings of a positive association between alpha-linolenic acid and a negative association between the ratio of linoleic to alpha-linolenic acid and the risk of prostate cancer.