Depression is a comorbid condition in Alzheimer's disease (AD) with negative consequences in patients and caregivers. Pathophysiology and optimal treatment are matters to be elucidated. A search of articles dealing with depression in AD was conducted in MEDLINE with special attention to epidemiology, pathophysiology, and treatment. Depression may predate dementia and tends to occur in up to 50% of AD patients with a decrease of noradrenalin and serotonin in the brain being the most plausible cause. Only 7 small double-blind randomized placebo-controlled clinical trials with antidepressants in AD patients with depression were found: 4 with sertraline, 1 with fluoxetine, 1 with imipramine, and another one with clomipramine. The total number of treated patients was 318. The weighted odds ratio (OR) was calculated with the method of Mantel-Haenszel. Both tricyclic antidepressants and selective serotonin reuptake inhibitors are better than placebo in treating depression in AD (weighted OR: 1.82, 95% CI: 1.13-2.96), with sertraline being one of the most used drugs. The differences were significant in 2 trials and not significant in four. The magnitude of effect is globally modest. Moreover, it is noteworthy mentioning the high rates of response to placebo in most studies. Depression is one of the most frequent behavioral symptoms in AD. Although antidepressants may work in AD, given the small number of patients treated, the effect is unclear. Further large randomized controlled clinical trials are warranted in order to know the best drug to begin with and the actual degree of efficacy.