To investigate the long-term influence of insulin resistance and hyperinsulinemia on vascular reactivity, both muscarinic and alpha (2)-receptor-mediated relaxations and the contribution of nitric oxide to these mechanisms were studied in the fructose-fed rat. Male Sprague-Dawley rats were fed either fructose-rich chow (FFR, n = 6) or normal chow (CNT. n = 6) for 40 weeks. Systolic blood pressure was measured by tail-cuff method. A 3-mm segment of mesenteric artery was excised, cannulated and pressurized, pretreated with prazosin (10(-6) mol/L) and propranolol (3 x 10(-6) mol/L), then precontracted with serotonin (10-6 mol/L). Endothelium dependent relaxation was induced by addition of acetylcholine (10(-9) to 10(-4) mol/L), or a selective alpha (2)-agonist B-HT 920 (10(-9) to 10(-5) mol/L), with or without the nitric oxide synthase inhibitor L-NAME (10(-4) mol/L). Systolic blood pressure was significantly higher in FFR at the early period; however, there was no difference at the end of 40 weeks compared to CNT. Fasting plasma insulin was much higher in FFR than in CNT (110 +/- 62 v 41 +/- 11 muU/mL, P < .05), whereas plasma glucose was not different. Maximum relaxation to acetylcholine was attained at 10(-6) mol/L in FFR but at 3 x 10(-7) mol/L in CNT. The degree of maximum relaxation attained with acetylcholine was similar in FFR and CNT (89 +/- 9 and 94 +/- 4% of precontraction), although attenuated (P < .01) by the addition of L-NAME only in FFR (to 34 +/- 22%, P < .05) but not in CNT (to 82 +/- 25%). The half-maximal relaxation dose of acetylcholine was greater in FFR (P < .01) compared with CNT and was significantly increased (P < .05) by L-NAME in both groups. B-HT 920 at 10(-5) mol/L induced a greater relaxation in CNT (36 +/- 10% of serotonin constriction) than in FFR (19 +/- 14%, P < .05). These responses were significantly blunted by L-NAME. Thus, muscarinic receptor-mediated vascular relaxation is less sensitive and more nitric oxide dependent in FFR versus CNT. alpha (2)-Adrenergic-mediated relaxation, predominantly mediated by nitric oxide, is also impaired in FFR. It is possible that prolonged insulin resistance and hyperinsulinemia in FFR could alter endothelial-dependent vasodilatory mechanisms, thereby contributing to the increase in blood pressure seen in this model. (C) 2001 American Journal of Hypertension. Ltd.