We studied the effects of apolipoprotein E (APOE) genotype and gender on clinical response to tacrine in patients with mild to moderate Alzheimer's disease (AD). We analyzed data from a previously reported 30-week, double-blind, placebo-controlled trial of tacrine, in which APOE genotypes were determined from previously collected plasma samples. Patients were assigned to placebo or tacrine with daily dosages of 80, 120, or 160 mg/day. The outcome measures were Alzheimer's Disease Assessment Scale-Cognitive Component, Clinician Interview Based Impression, Mini-Mental State Examination, and the Caregiver-rated Clinical Global Impression of Change. An intent-to-treat (ITT) analysis of patients with available genotypes (n = 528) did not reveal response differences by genotype, although the effect size was twice as large in the epsilon2-3 as the epsilon4 group (-2.62 versus -1.25). The association of treatment effect with APOE genotype varied significantly according to gender (p < 0.002 for ITT; p < 0.05 for evaluables). The treatment effect was larger in the epsilon2-3 compared with epsilon4 women (ITT, 4.24 points, p = 0.03; evaluable, 7.20 points, p = 0.01). In contrast, treatment effect size was not different between epsilon2-3 and epsilon4 of men with AD. APOE genotype and gender may predict response to tacrine in patients with AD.