Abnormalities in the metabolism of essential fatty acids and the results of increased oxidative stress have been implicated in cardiovascular disorders observed in diabetes mellitus. This study, therefore, aimed to investigate the effects of cod liver oil (CLO, Lysi Ltd, Iceland), which comprises mainly an antioxidant vitamin A, n:3 polyunsaturated fatty acids (n:3 PUFAs), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), on cardiovascular abnormalities in streptozotocin (STZ)-diabetic rats. Two days after single STZ (55 mg kg(-1), i.p.) or vehicle injection, diabetes was verified by increased blood glucose, and non-diabetic and diabetic rats were left untreated or treated with CLO (0.5 mL kg(-1) daily, by intragastric probing) for 12 weeks. Plasma glucose, triacylglycerol and cholesterol concentrations were significantly elevated in 12-week untreated-diabetic rats; CLO provided better weight gain, entirely prevented the plasma lipid abnormalities, but partially controlled the glycaemia in diabetic rats. In isolated aorta rings, diabetes resulted in increased phenylephrine-induced vasoconstriction and isoprenaline-induced vasorelaxation, impaired endothelium-dependent vasodilatation and unchanged responsiveness to sodium nitroprusside. CLO treatment completely prevented endothelial deficiency, partly corrected the phenylephrine-induced vasoconstriction and did not affect the responses to isoprenaline and sodium nitroprusside in diabetic aorta. Diabetes also produced a marked decrease in the rate of spontaneously beating right atria and a significant increase in basal contractile force of left ventricular papillary muscle. The responsiveness of right atria to the positive chronotropic effect of isoprenaline was significantly decreased in diabetic rats, and was increased in CLO-treated diabetic rats. The positive chronotropic effect of noradrenaline was markedly increased in diabetic atria, but prevented by CLO treatment. Diabetes also resulted in an increased positive inotropic response of papillary muscle to both noradrenaline and isoprenaline, which were prevented by CLO treatment. CLO treatment also resulted in lower tissue sensitivity (pD(2)) to these agonists in diabetic papillary muscle. Ventricular hydroxyproline content was found to be unchanged among the experimental groups. The ultrastructure of diabetic myocardium displayed various degenerations (i.e. intracellular oedema, myofibrillar fragmentation, condensed pleomorphic mitochondria, thick capillary irregular basement membrane, swollen endothelial cells), which were partially prevented by CLO treatment. We conclude that the supplementation with CLO is effective in preventing cardiovascular disorders observed in experimental diabetes.