This review addresses the question of whether there is evidence that antidepressants are more efficacious than placebo in the treatment of late-life depression and what is the rate of response that physician and patient can expect when antidepressant medication is prescribed in a typical clinical setting. To date, 5 placebo-controlled and 10 comparison trials have study designs of sufficient rigor to provide evidence of antidepressant efficacy and effectiveness in the treatment of late-life depression. The results suggest that antidepressant medications are more effective than placebo. However, placebo-controlled trials are not a simple comparison of only medication versus placebo. Rather, the amount of nonspecific psychosocial interventions included in these trials is considerable and often not systematically measured. Trial design also affects outcome: response and remission rates in comparison trials consistently are 20% to 30% higher than those reported in placebo-controlled trials. Clinical trials do not consistently assess the many moderators that are believed to affect treatment outcome in late-life depression, and therefore, comparisons across studies are problematic because of an inability to determine whether patient samples are truly comparable. For future clinical trials to have maximal relevance, study design should evolve to reflect as closely as possible a typical clinical setting especially with respect to frequency and duration of patient visits.