The dependence of blood pressure on a balance between superoxide and nitric oxide may be amplified in diabetes. We have shown that the first occurrence of sustained hyperglycemia in type I diabetes causes hypertension when induced in rats that have had nitric oxide synthesis blocked chronically (L-NAME, 10 microg/kg per minute IV). This study used tempol (18 micromol/kg per hour IV) to test the hypothesis that superoxide mediates that hypertensive response. Induction of diabetes in untreated rats had no significant effect on mean arterial pressure (MAP, measured 18 h/d), and glomerular filtration rate (GFR) increased significantly during the 2 weeks of diabetes. Chronic infusion of L-NAME in a separate group of rats increased baseline MAP from approximately 90 mm Hg to a stable level of approximately 120 mm Hg after 6 days of infusion, and induction of diabetes (streptozotocin, 40 mg/kg IV) in those rats caused a rapid, progressive increase in MAP that averaged 156+/-5 mm Hg by day 14 of diabetes that was associated with a decrease in GFR and 4-fold increase in isoprostane excretion. Tempol infusion was begun on day 2 of diabetes in a subgroup of those rats, and the progressive hypertensive response was prevented, with MAP averaging 134+/-10 mm Hg by day 14. In addition, the normal renal hyperfiltration response was restored by tempol and the increase in isoprostane did not occur. Thus, the hypertension and decrease in GFR caused by onset of diabetes in rats without a functioning nitric oxide system was prevented by chronic administration of the superoxide dismutase mimetic tempol.