We examined microvascular responses to acetylcholine (ACh), adenosine (ADO), nitroprusside (SNP), bradykinin (BK), histamine (HIS), and serotonin (5-HT) in control and diabetic rats. Agonists were applied topically to neurovascularly intact and environmentally controlled cremaster muscles of diabetic or control rats 3 weeks after streptozotocin or vehicle injection. Precapillary arteriolar diameters and leakage of fluorescein isothiocyanate conjugated to bovine serum albumin (FITC-BSA) in postcapillary venules were measured by using intravital microscopy techniques. All agents dilated the arterioles, and BK, HIS, and 5-HT caused concentration-dependent leakage of FITC-BSA. Of the vasodilators tested, only ACh-induced dilation was attenuated in diabetic animals. BK and HIS caused a concentration-dependent leakage of FITC-BSA that was similar in diabetic and control groups, but FITC-BSA leakage to 5-HT was significantly attenuated in diabetic animals. To determine whether the attenuated responses were attributable to an alteration of nitric oxide, the same dilational and leakage-promoting agents were tested in normoglycemic animals in the presence of N'-nitro-L-arginine methyl ester (L-NAME), an inhibitor of nitric oxide synthase. Only ACh-induced dilation in control animals was attenuated in the presence of L-NAME; all other responses were normal. Thus, although the attenuated ACh-induced arteriolar dilation in striated muscle of diabetic rats may be linked to an impairment of nitric oxide function or release, the reduced leakage to serotonin is not linked to nitric oxide. The impaired leakage response to 5-HT is specific for that agonist and is not an indication of a generalized decrease in vascular permeability in these diabetic animals.