Cardiovascular effects of an essential oil from the aerial parts of Mentha x villosa (OEMV) were tested in rats using a combined in vivo and in vitro approach. In non-anesthetized normotensive rats, OEMV (1, 5, 10, 20, 30 mg kg(-1) body wt., i.v.) induced a significant and dose-dependent hypotension (-3 +/- 1.8%; -6 +/- 0.7%; -40 +/- 6.7%; -58 +/- 3.8%; -57 +/- 2.1%, respectively) associated with decreases in heart rate (-1 +/- 0.3%; -9 +/- 0.9%; -17 +/- 3.2%; -72 +/- 3.1%; -82 +/- 1.4%, respectively). The hypotensive and bradycardic responses evoked by OEMV were attenuated and blocke by pre-treatment of the animals with atropine (2 mg kg(-1) body wt., i.v.). In isolated rat atrial preparations, OEMV (10, 100, 300, 500 microg ml(-1)) produced concentration-related negative chronotropic and inotropic effects (IC50 value = 229 +/- 17 and 120 +/- 13 microg ml(-1), respectively). In isolated rat aortic rings, increasing concentrations of OEM (10, 100, 300, 500 microg ml(-1)) were able to antagonize the effects of phenylephrine (1 microM), prostaglandin F2alpha (10 microM) and KCl (80 mM)-induced contractions (IC50 value = 255 +/- 9, 174 +/- 4 and 165 +/- 14 microg ml(-1), respectively). The vasorelaxant activity induced by OEMV was attenuated significantly by either endothelium removal (IC50 value = 304 +/- 9 microg ml(-1)), NG-nitro L-arginine methyl ester (L-NAME) 100 microM (IC50 value=359 +/- 18 microg ml(-1)), L-NAME 300 microM (IC50 value = 488 +/- 20 microg ml(-1)) or indomethacin 10 microM (IC50 value = 334 +/- 18 microg ml(-1)). However, it was not affected by atropine 1 microM (IC50 value = 247 +/- 12 microg ml(-1)). Furthermore, the hypotensive response induced by OEMV was attenuated significantly after nitric oxide (NO) synthase blockade (L-NAME, 20 mg kg(-1) body wt., i.v.), while bradycardia was not altered. The results suggest that the hypotensive effect induced by OEMV is probably due to its direct cardiodepressant action and peripheral vasodilation, which can be attributed to both endothelium-dependent (via EDRFs, at least NO and prostacyclin) and endothelium-independent mechanisms (such as Ca2+ channel blockade).