These studies were undertaken to investigate the relationship between regional hemodynamic and hemorheological changes in the microvasculature of diabetic rats. Diabetes was induced in male Sprague-Dawley rats by injection of streptozotocin (55 mg/kg body wt). Control rats were injected with vehicle (sodium citrate buffer). A subgroup of diabetic rats was treated with an aldose reductase inhibitor (sorbinil) added to the diet in an amount to provide a daily dose of almost-equal-to 0.2 mmol.kg-1.day-1. Three weeks later all animals were anesthetized with thiobutabarbital sodium (Inactin, 100 mg/kg injected intraperitoneally) for assessment of blood flow (by injection of 15 mum microspheres) and regional hematocrit (determined by isotope-dilution techniques using Cr-51-labeled red blood cells and I-125-labeled bovine serum albumin) in selected tissues. The hematocrit in arterial blood samples was identical (approximately 46%) in controls and in diabetics. Regional hematocrits were much lower than arterial hematocrits in control rats and ranged from approximately 20% in ocular tissues, sciatic nerve, diaphragm, and skin to approximately 30% in brain, skeletal muscle, heart, and fat. Hematocrits of diabetic rats were markedly increased in ocular tissues, sciatic nerve, and skin but not in brain, heart, or skeletal muscle. These increases in regional hematocrit were associated with increases in blood flow and were largely prevented by sorbinil. Diabetes induced significant decreases in the mean transit times for whole blood and erythrocytes in all tissues examined except brain, retina, and skin. The demonstration and correspondence of selective increases in blood flow and hematocrit in the same tissues suggest the possibility that an interaction between diabetes-induced hemodynamic and hemorheological changes may contribute to the pathogenesis of microangiopathy in such tissues and may account for their selective predilection, independent of systemic hemorheological alterations, as sites of clinically significant microangiopathy in human diabetics.