Objectives: To evaluate the safety of the peroxisome proliferator-activated receptor gamma agonist pioglitazone in nondiabetic patients with Alzheimer disease (AD) and to explore treatment effect sizes on clinical outcomes. Design: Double-blind, placebo-controlled randomized controlled trial of 18-month duration. Setting: Two academic medical center outpatient clinics. Patients: Nondiabetic patients meeting research criteria for probable AD were enrolled. Twenty-five of 29 subjects completed the study; no withdrawals were attributable to adverse effects. Intervention: Subjects received pioglitazone (Actos), titrated to 45 mg daily, or matching placebo, and 200 IU of vitamin E daily. Patients maintained treatment with cholinesterase inhibitors and could begin memantine therapy when it became available during the study. Main Outcome Measures: The primary outcome was frequency of reported adverse effects (AEs). Secondary outcomes were measures of cognition, activities of daily living, neuropsychiatric symptoms, and global function. Results: Peripheral edema was the principal AE occurring more frequently in subjects taking pioglitazone than placebo (28.6% vs 0%). This is consistent with the known AE profile of pioglitazone. No group differences in laboratory measures were identified. No significant treatment effect was observed on exploratory analysis of clinical efficacy. Conclusions: Pioglitazone was generally well tolerated in this pilot study. There were no serious or unanticipated adverse events or clinical laboratory changes attributable to pioglitazone over a long-term exposure in nondiabetic patients with AD. The tolerability of pioglitazone in this population and peroxisome proliferator-activated receptor gamma effects in laboratory models of AD support further study of this drug class in earlier disease stages. Trial Registration: clinicaltrials.gov Identifier: NCT00982202.