The present study was designed to test whether altered renovascular reactivity is associated with the increased risk of radio-contrast nephropathy (RCN) in diabetics. We studied 50 patients (24 diabetics, 26 nondiabetics) with chronic renal insufficiency undergoing cardiac catheterization. Patients were randomized to receive either saline, or one of three renal vasodilator/diuretic drugs--dopamine, atrial natriuretic peptide (ANP), or mannitol--by intravenous infusion during cardiac catheterization. Renal blood flow (RBF) was measured by thermodilution at various time points during cardiac catheterization. RCN was defined as an increase in PCr of at least 25% over baseline within 48 hours of cardiac catheterization. Baseline PCr and creatinine clearance were similar in diabetics and nondiabetics (2.6 +/- 0.2 mg/dl vs. 2.4 +/- 0.1 mg/dl, and 32 +/- 3 ml/min vs. 34 +/- 3 ml/min, respectively), but baseline RBF was significantly lower in diabetics (154 +/- 21 ml/min/kidney vs. 277 +/- 36 ml/min/kidney, P < 0.05). Diabetic patients exposed to the three vasodilator/diuretic drugs had the greatest increase in RBF throughout cardiac catheterization. The incidence of RCN among the diabetics receiving those drugs was 83%, 83% and 75%, in the dopamine, ANP and mannitol groups, respectively. In contrast, among the nondiabetics in each of those groups the incidence of RCN was zero (all P < 0.05, diabetics vs. nondiabetics). In the saline control group the rates of RCN in the diabetics and nondiabetics were 43% and 38%, respectively (NS). In conclusion, the increased risk of RCN among diabetics was associated with exaggerated renovascular reactivity: baseline vasoconstriction and enhanced vasodilation with vasodilator/diuretic drugs. These same drugs, however, reduced the risk of RCN in nondiabetic patients.