Objective: To estimate the incidence rate and predictors of seizures in patients with mild to moderate Alzheimer disease (AD). Design: Cohort study of patients with mild to moderate AD in clinical trials. Risk factors for potential seizures were evaluated by stratified descriptive statistics and univariable and multivariable Cox proportional hazards regressions. Setting: Pooled patient-level data from 10 Alzheimer Disease Cooperative Study clinical trials in mild to moderate AD from 1995 to 2010. Patients: Three thousand seventy-eight subjects randomized to the treatment or placebo arms of 10 AD clinical trials. Screening Mini-Mental State Examination scores ranged between 10 and 28. Results: Eighteen seizures were reported in 3078 randomized subjects, with an incidence rate of 484 per 100 000 person-years (95% CI, 287-764). Statistically significant independent risk factors for seizure were younger age (adjusted hazard ratio, 0.80; 95% CI, 0.69-0.93 per every 5 years of age), greater cognitive impairment at baseline (adjusted hazard ratio, 2.79; 95% CI, 1.06-7.33 for Mini-Mental State Examination scores <18 compared with Mini-Mental State Examination scores ≥18), and antipsychotic use at baseline (adjusted hazard ratio, 3.47; 95% CI, 1.33-9.08). Conclusions: Seizure rates in patients with mild to moderate AD in clinical trials are similar to rates observed in longer observational cohort studies, but they are greater than expected in the general elderly population. Younger age, greater degree of cognitive impairment, and history of antipsychotic use were independent risk factors for new-onset seizures in AD.