Fish oil polyunsaturated fatty acids and fibrate hypolipidemic drugs are potent hypotriglyceridemic agents that act by increasing fatty acid catabolism and decreasing triglyceride synthesis and secretion by the liver. A major unresolved issue is whether this hypotriglyceridemic effect can occur independent of induction of peroxisomal beta-oxidation, a predisposing factor for hepatocarcinogenesis. The present study was undertaken to determine which component of fish oil, eicosapentaenoic acid (EPA) or docosahexaenoic acid (DHA), is responsible for its triglyceride-lowering effect. We demonstrate that EPA and not DHA is the hypotriglyceridemic component of fish oil and that mitochondria and not peroxisomes are the principal target. Results obtained by fenofibrate feeding support the hypothesis that the mitochondrion is the primary site for the hypotriglyceridemic effect. In contrast to fibrates, EPA did not affect hepatic apolipoprotein C-III gene expression. Therefore, increased mitochondrial beta-oxidation with a concomitant decrease in triglyceride synthesis and secretion seems to be the primary mechanism underlying the hypotriglyceridemic effect of EPA and fibrates in rats, rabbits and possibly also in humans. In addition, these data show that lowering of plasma triglycerides can occur independently of any deleterious peroxisome proliferation.