This study examined the efficacy and safety of memantine monotherapy in patients with moderate-to-severe Alzheimer disease (AD). Patients not receiving a cholinesterase inhibitor (N=350) were randomized to receive memantine (20 mg/d) or placebo during this 24-week, double-blind, placebo-controlled trial. Prospectively defined analyses failed to demonstrate a statistically significant benefit of memantine treatment compared with placebo on the Severe Impairment Battery (SIB) at week 24 end point, although a significant advantage was observed for memantine at weeks 12 and 18. The 19-item Alzheimer's Disease Cooperative Study-Activities of Daily Living Scale (ADCS-ADL₁₉) did not differ significantly between groups in any analysis. Clinician's Interview-Based Impression of Change plus Caregiver Input (CIBIC-Plus) did not significantly favor memantine at week 24 despite a significant advantage for memantine at weeks 12 and 18. Other secondary outcomes showed no significant treatment differences. Post hoc analyses of potentially confounding covariates and alternative methods of imputing missing data did not substantially alter the results. Because of the violations of normality assumptions for the SIB and ADCS-ADL₁₉, nonparametric analyses were performed; statistically significant benefit of memantine over placebo was demonstrated at week 24 for the SIB but not the ADCS-ADL₁₉. The type and incidence of adverse events were similar in both groups.