Major depression afflicts approximately 25 percent of patients with AD. Depression in AD patients leads to mental suffering, behavioral disturbance (such as aggression), poor cognition, poor self-care, caregiver depression, caregiver burden, and early entry into the nursing home. Since major depression is treatable, this additional disability may be avoidable. The use of antidepressants to treat major depression in AD is supported by two studies, although a third does not show a benefit for antidepressants over placebo. Also, the safety of antidepressant treatment in depressed AD patients is poorly studied. A conclusive study showing that depression reduction in AD can be accomplished safely with antidepressant medications, and that depression reduction is associated with improvements in activities of daily living, non-mood behavioral disturbances, caregiver burden, and caregiver depression would have major clinical and cost implications for the care of the AD patient. This study is a 13-week, double blind, flexible dose, placebo controlled trial of sertraline in the treatment of outpatients with AD and co-morbid major depression. The hypothesis is that antidepressant treatment is superior to placebo in improving mood, in improving cognition, in reducing physical dependency, in reducing caregiver depression, and in reducing caregiver burden. It is also hypothesized that the degree of depression reduction is correlated with these improvements. It is further hypothesized that the safety profile of sertraline when compared to placebo is acceptable, especially with regard to risk of falls, sleep disturbance, and delirium. One hundred community residing outpatients with probable AD who also meet DSM-IV criteria for major depressive episode will be recruited into the study. They will be randomized to sertraline or placebo and followed through weekly telephone contact by an experienced clinical trials team. Outcomes will be assessed every 3 weeks, for a total of four followup data points. Scales assessing the following domains will be used: depression, cognition, behavioral disturbance, physical dependency, delirium, falls, sleep, other side effects, caregiver depression, caregiver burden, caregiver functioning, and caregiver health.