The isolated perfused rabbit ear was used to investigate the effect of eicosapentaenoic acid (EPA) on the vasoconstriction induced by noradrenaline (NA) and angiotensin II (A II). EPA (0.1, 1.0, 10.0 micrograms/ml) was shown to reduce the vasoconstriction (inhibition of flow) induced by NA in a dose-dependent manner. The dose-response curves of NA were shifted to the right. Emax of NA was reduced in particular at the highest concentration of EPA. The maximum effect of EPA developed slowly and reached a steady state after 150 min (0.1 microgram/ml EPA), 120 min (1.0 microgram/ml EPA), 100 min (3.0 microgram/ml EPA) and 60 min (10 micrograms/ml), respectively. In the presence of indomethacin (3 micrograms/ml), the vasoconstrictor effect of NA was not significantly influenced. The effect of EPA (3 micrograms/ml) was not reduced by indomethacin. In the presence of the lipoxygenase inhibitor nordihydroguaiaretic acid (NDGA, 10 micrograms/ml), the vasoconstrictor effect of NA was not significantly influenced. The effect of EPA (3 micrograms/ml) was not reduced by NDGA. Concomitant addition of indomethacin (3 micrograms/ml) and NDGA (10 micrograms/ml) also did not reduce the effect of EPA. EPA (3.0, 10.0 micrograms/ml) was shown to reduce also the vasoconstrictor effect of A II in the absence and the presence of the cyclooxygenase and lipoxygenase inhibitors. Determination by radioimmunoassay showed that the levels of thromboxane B2 (TXB2)-like immunoreactivity were below the detection limit in the perfusate of the ear. The levels of 6-keto-PGF1 alpha- and PGE2-like immunoreactivity in the perfusate were low. EPA reduced the concentration of 6-keto-PGF 1 alpha.(ABSTRACT TRUNCATED AT 250 WORDS)