Human obesity is associated with insulin resistance, hyperinsulinemia, and a predisposition to hypertension and vascular disease, the origin of which may lie in impairment of endothelial function. We tested the effects of the thiazolidinedione rosiglitazone on blood pressure and endothelial function in insulin-resistant fatty Zucker rats, which display hypertension and abnormal endothelial cell function. We studied fatty Zucker rats given rosiglitazone maleate (50 micromol/kg diet; n = 8) for 9-12 weeks (treated fatty), untreated fatty rats (n = 8), and lean rats (n = 8) given diet alone. At the end of the study, systolic blood pressure was significantly higher in untreated fatty (147 +/- 5 mmHg) than in lean rats (125 +/- 2 mmHg; P < 0.05), but rosiglitazone treatment prevented the development of hypertension in fatty rats (123 +/- 1 mmHg). Fasting hyperinsulinemia in untreated fatty rats (28.7 +/- 6.0 ng/ml) was significantly lowered by rosiglitazone (7.0 +/- 1.4 ng/ml; P < 0.05 vs. untreated fatty), but remained significantly higher than the levels seen in lean rats (1.5 +/- 0.4 ng/ml; P < 0.01). Mesenteric arteries were studied in a myograph. Maximal acetylcholine chloride (1.1 micromol/l)-induced relaxation of norepinephrine hydrochloride (NE)-induced constriction was impaired in untreated fatty (62.4 +/- 3.4%) vs. lean (74.3 +/- 3.5%; P = 0.01) rats; this defect was partially prevented by rosiglitazone (66.5 +/- 3.0%; P = 0.01 vs. untreated fatty). Insulin (50 mU/l) significantly attenuated the contractile response to NE in lean rats (14.7 +/- 3.3%; P = 0.02); this vasodilator effect of insulin was absent in untreated fatty rats at concentrations of 50-5,000 mU/l, but was partially restored by rosiglitazone (9.7 +/- 2.5% attenuation; P = 0.02 vs. no insulin). Thus, rosiglitazone prevents the development of hypertension and partially protects against impaired endothelial function associated with insulin resistance. These latter effects may contribute to the drug's antihypertensive properties.