The efficacy of ifosfamide (IFO), an antineoplastic drug, is severely limited by a high incidence of nephrotoxicity of unknown etiology. We hypothesized that inhibition of complex I (C-I) by chloroacetaldehyde (CAA), a metabolite of IFO, is the chief cause of nephrotoxicity, and that agmatine (AGM), which we found to augment mitochondrial oxidative phosphorylation and beta-oxidation, would prevent nephrotoxicity. Our model system was isolated mitochondria obtained from the kidney cortex of rats treated with IFO or IFO + AGM. Oxidative phosphorylation was determined with electron donors specific to complexes I, II, III, or IV (C-I, C-II, C-III, or C-IV, respectively). A parallel study was done with (13)C-labeled pyruvate to assess metabolic dysfunction. Ifosfamide treatment significantly inhibited oxidative phosphorylation with only C-I substrates. Inhibition of C-I was associated with a significant elevation of [NADH], depletion of [NAD], and decreased flux through pyruvate dehydrogenase and the TCA cycle. However, administration of AGM with IFO increased [cyclic AMP (cAMP)] and prevented IFO-induced inhibition of C-I. In vitro studies with various metabolites of IFO showed that only CAA inhibited C-I, even with supplementation with 2-mercaptoethane sulfonic acid. Following IFO treatment daily for 5 days with 50 mg/kg, the level of CAA in the renal cortex was approximately 15 micromol/L. Taken together, these observations support the hypothesis that CAA is accumulated in renal cortex and is responsible for nephrotoxicity. AGM may be protective by increasing tissue [cAMP], which phosphorylates NADH:oxidoreductase. The current findings may have an important implication for the prevention of IFO-induced nephrotoxicity and/or mitochondrial diseases secondary to defective C-I.