To evaluate the effect of ethyl icosapentaenoate (EPA) on the metabolism of tumor necrosis factor alpha (TNFalpha) and interleukin-1beta (IL-1beta), the concentrations of these cytokines in the carotids of rabbits sheathed in a cuff were studied. Japanese white rabbits were divided into two groups; the EPA group, in which 600 mg/kg/day EPA was administered forcibly p.o. for 1 week before cuff treatment, and the control group. Carotid artery samples were obtained just before, 3 days and 7 days after cuff treatment, and TNFalpha and IL-1beta were determined separately with the Western blot analysis method. In the control group, there were 43.5 (+/- 3.0) pg/microg protein of TNFalpha and 53.5 (+/- 4.8) pg/microg protein of IL-1beta just before cuff treatment. Compared to the control group, these concentrations of the EPA group were both significantly low. Three days after cuff treatment, TNFalpha of the EPA group was still significantly low, while IL-1beta showed no difference. There was no significant difference between the two groups 7 days after cuff treatment. These findings suggested that EPA could influence TNFalpha and IL-1beta metabolism in the arterial wall even at baseline. Furthermore, EPA suppressed TNFalpha and IL-1beta production in the early phase of intimal thickening, indicating a mechanism inhibiting the activation of smooth muscle cells such as their proliferation and migration, induced by the cuff-sheath method.