Naturally occurring hydroxystilbenes have been shown to induce vasorelaxation. Here, we studied the mechanism of resveratrol-induced vasorelaxation in different types of blood vessels, namely mesenteric (resistance) and main uterine (conductance) arteries, from female guinea-pigs on day 7 and day 15 of the oestrous cycle. Resveratrol (5-70 micromol/l) induced concentration-dependent relaxation of both mesenteric and uterine arteries preconstricted with either noradrenaline (NA; 10 micromol/l) or KCl (125 mmol/l). Resveratrol was 2-fold more potent in inducing relaxation of mesenteric arteries than of uterine arteries. Its effects on uterine arteries from both day-7 and day-15 guinea-pigs were similar, irrespective of the constrictor used, but it was significantly (P<0.01) more potent in inducing relaxation of mesenteric arteries contracted with NA compared with those constricted with KCl. In day-7 arteries precontracted with NA, N(G)-nitro-L-arginine methyl ester (L-NAME; 10 micromol/l) had no effects on the time course of resveratrol-induced vasorelaxation in either mesenteric or uterine arteries. However, indomethacin (50 micromol/l) significantly (P<0.05) potentiated resveratrol's effect on mesenteric, but not uterine, arteries. Indomethacin had no effect on resveratrol-induced vasorelaxation of arteries contracted with KCl, whereas L-NAME significantly (P<0.05) reduced the effects of resveratrol on uterine, but not on mesenteric, arteries. In day-15 arteries, L-NAME significantly (P<0.01) attenuated the effects of resveratrol on mesenteric arteries contracted with NA. Indomethacin had no effect on resveratrol activity. This study indicates that: (a) the effect of resveratrol on resistance arteries is greater than that on conductance arteries; (b) the effects of resveratrol are not mediated via prostanoids, but NO may play a role; and (c) the stage of the oestrous cycle has no influence on resveratrol-induced vasorelaxation.