Escitalopram is a selective serotonin reuptake inhibitor (SSRI); it is the therapeutically active S-enantiomer of the racemic mixture, citalopram. This review aimed to compare the efficacy and tolerability of escitalopram versus citalopram and several other SSRIs (citalopram, fluoxetine, paroxetine, sertraline), and a selective reuptake inhibitor of noradrenaline and serotonin, venlafaxine XR, for treatment of DSM IV (Diagnostic and Statistical Manual of mental disorders - fourth edition) major depressive disorder, based on the studies evaluated by the Commission de la Transparence de la R6publique Frangaise, and data from a pooled analysis presented in 2005 at the 158th annual congress of the American Psychiatric Association. Change from baseline to end-point on total MADRS (Montgomery-Asberg Depression Rating Scale--10 items, score range: 0-60) was the primary efficacy parameter; changes on HAM-D17 (Hamilton rating scale for depression--17 items), CGI-S and CGI-I (Clinical global Impression-Severity and-Improvement), and response rates (> or = 50% MADRS score reduction) and remissions (< 12 MADRS score) were the secondary efficacy parameters. Tolerability assessment was based on the numbers and rates of adverse events observed with treatment, and the DESS (Discontinuation Emergent Signs and Symptoms-43 items) scale was used for assessment of adverse events observed with treatment withdrawal. Analyses were based on intention to treat using the LOCF (last observation carried forward) method. Efficacy of escitalopram appeared to be at least equivalent to that of the active comparators in all cases. The difference between active compounds was more marked when depressive symptoms were more severe. From the point of view of tolerability, frequency of adverse effects occurring on treatment and the frequency of treatment discontinuations due to adverse effects were comparable with both escitalopram and the active comparators; however, the comparisons were mostly favourable to escitalopram, though differences were generally not statistically significant. In both studies of escitalopram versus venlafaxine XR, treatment discontinuations due to adverse events were less frequent on escitalopram than on venlafaxine XR (7.5% vs 11.2%, and 4.1% vs 16.0% respectively). With regard to adverse events associated with the withdrawal period, the signs and symptoms occurring on treatment discontinuation assessed after 1 week using the DESS scale were less frequent on escitalopram than on venlafaxine XR at 8 weeks and paroxetine at 24 weeks. Concerning suicide risk, a review of clinical trials involving 2277 patients on escitalopram and 1814 patients on placebo showed that this risk was minimal, and similar in both groups; moreover, no evidence was found suggesting that escitalopram might promote suicidal behaviour compared with placebo. These results suggest that escitalopram is suitable to be considered as a first-line drug treatment for major depressive disorder.