1 We have determined the dermal microvascular effects of the nitric oxide (NO) synthase inhibitor, NG-nitro-L-arginine methyl ester (L-NAME, 100 nmol/site), endothelin-1 (ET-1, 0.1-10 pmol/site) and ET-3 (0.1-30 pmol/site) in rats with streptozotocin (STZ)-induced diabetes mellitus. Cutaneous blood flow changes as measured by a 133xenon (133Xe) clearance technique, were determined in diabetic rats four weeks after treatment with streptozotocin (STZ) and compared with responses measured in normal rats four weeks after treatment with saline. 2 Resting skin blood flow was similar in diabetic and in normal rats, as measured by 133Xe clearance and laser Doppler flowmetry. 3 Intradermal NG-nitro-L-arginine methyl ester (L-NAME) reduced skin blood flow in normal rats by 55.2 +/- 2.6% as measured by 133Xe clearance, (n = 9). L-NAME was significantly less effective in diabetic rats, inducing a 40.9 +/- 7.7% decrease in blood flow (n = 9, P less than 0.05). The enantiomer D-NAME had no effect in either group of rats. 4 Low doses of ET-1 and ET-3 injected intradermally induced dose-dependent decreases in blood flow, measured by 133Xe clearance, which were similar in both groups of rats. However, the responses to the highest doses of ET-1 (10 pmol/site) and ET-3 (10 and 30 pmol/site) were significantly reduced in the diabetic compared with the normal rats (P less than 0.05).In addition vasoconstriction to the highest doses of vasopressin (0.3 and 3 pmol/site) and vasodilatation to the neuropeptide calcitonin gene-related peptide (CGRP, 1O pmol/site) were similarly reduced in the diabetic rats (P <0.05).5. The decrease in blood flow induced by submaximal doses of ET-1 was enhanced by co-injection with L-NAME (100 nmol/site) in both diabetic and normal rats. However, this enhanced response was significantly reduced in the diabetic rats (P<0.05). A similar pattern of responses were observed to ET-3 in the presence and absence of L-NAME.6. These results indicate that the cutaneous microvasculature of rats with STZ-induced diabetes responds differently to intradermal ET-1 and ET-3 compared with normal rats; a similarly altered vascular reactivity was observed with vasopressin and CGRP. Hence, the diabetic microcirculation has impaired responses to several vasoconstrictors and a vasodilator. The effect of the nitric oxide synthase inhibitor L-NAME is also suppressed in the diabetics, suggesting that there may be decreased local production of, or response, to nitric oxide.