In human hypertension, blockade of beta-adrenoceptors does not improve resistance artery structure or endothelial dysfunction. We tested in hypertensive rats the hypothesis that carvedilol, a beta-blocker with antioxidant properties, would improve endothelial dysfunction, whereas the beta1-selective blocker, metoprolol, would not. Twenty-week-old SHRSP were treated orally for 10 weeks with carvedilol (50 mg/kg/day) or metoprolol (100 mg/kg/day), with or without hydralazine (25 mg/kg/day), the latter because neither beta-blocker was a very effective blood pressure-lowering agent in this model. Mesenteric arteries (lumen, <300 microm) were studied on a pressurized myograph. After 10 weeks, untreated SHRSP had a systolic blood pressure (mm Hg) of 239+/-3 that was unaffected by carvedilol or metoprolol treatment but decreased (p < 0.05) by hydralazine (187+/-4), carvedilol + hydralazine (221+/-3), and metoprolol + hydralazine (197+/-3). Carvedilol alone improved endothelium-dependent relaxation of resistance arteries, as elicited by the lowest concentration of acetylcholine studied (10(-7) M), whereas metoprolol had no effect. Hydralazine improved endothelial function as elicited by acetylcholine at a dose of 10(-6) M, also found under cotreatment with carvedilol but attenuated by cotreatment with metoprolol. Carvedilol or metoprolol alone had no significant effect on endothelium-independent relaxation produced by a nitric oxide donor (sodium nitroprusside). However, vessels from rats treated with carvedilol + hydralazine exhibited significantly greater relaxation than those from rats treated with metoprolol + hydralazine. These data suggest that carvedilol may have favorable effects on hypertension-related endothelial dysfunction not observed with metoprolol. Neither drug corrected small artery structure in SHRSP.