We tested the hypothesis that reactive oxygen species (ROS) contributed to renal hypoxia in C57BL/6 mice with ⅚ surgical reduction of renal mass (RRM). ROS can activate the mitochondrial uncoupling protein 2 (UCP-2) and increase O(2) usage. However, UCP-2 can be inactivated by glutathionylation. Mice were fed normal (NS)- or high-salt (HS) diets, and HS mice received the antioxidant drug tempol or vehicle for 3 mo. Since salt intake did not affect the tubular Na(+) transport per O(2) consumed (T(Na/)Q(O2)), further studies were confined to HS mice. RRM mice had increased excretion of 8-isoprostane F(2α) and H(2)O(2), renal expression of UCP-2 and renal O(2) extraction, and reduced T(Na/)Q(O2) (sham: 20 ± 2 vs. RRM: 10 ± 1 μmol/μmol; P < 0.05) and cortical Po(2) (sham: 43 ± 2, RRM: 29 ± 2 mmHg; P < 0.02). Tempol normalized all these parameters while further increasing compensatory renal growth and glomerular volume. RRM mice had preserved blood pressure, glomeruli, and patchy tubulointerstitial fibrosis. The patterns of protein expression in the renal cortex suggested that RRM kidneys had increased ROS from upregulated p22(phox), NOX-2, and -4 and that ROS-dependent increases in UCP-2 led to hypoxia that activated transforming growth factor-β whereas erythroid-related factor 2 (Nrf-2), glutathione peroxidase-1, and glutathione-S-transferase mu-1 were upregulated independently of ROS. We conclude that RRM activated distinct processes: a ROS-dependent activation of UCP-2 leading to inefficient renal O(2) usage and cortical hypoxia that was offset by Nrf-2-dependent glutathionylation. Thus hypoxia in RRM may be the outcome of NADPH oxidase-initiated ROS generation, leading to mitochondrial uncoupling counteracted by defense pathways coordinated by Nrf-2.