Severe forms of depression are a major therapeutic concern for psychiatrists. According to Health Authority recommendations, they require the systematic initiation of treatment with active drugs, and they respond less well to placebos than the less severe forms. Escitalopram, which is the most active enanthiomer of the racemic compound (citalopram), is tolerated well at the doses indicated in the marketing authorisation (10 to 20mg per day) and it is particularly effective in the severe forms of depression. Several studies have compared escitalopram to another specific serotonin recapture inhibitor (SSRI) in severe depression. In a 24-week study, 20mg per day of escitalopram, compared to 40mg per day paroxetin, demonstrated significantly greater efficacy (p<0.05) on the primary criterion (modification of the total MADRS score between inclusion and the end of the study). In this same study, the difference in favour of escitalopram increased in parallel with the increase in initial severity. In a grouped analysis of three studies versus citalopram, the superior efficacy of escitalopram also increased in parallel with the initial severity. The antidepressants, combined serotonin recapture inhibitors and noradrenalin (SRINA), might, in theory, be more effective than the SRI because of their broader mode of action. Recent data on escitalopram have invalidated this fact. In a study comparing 20mg per day of escitalopram to 225mg per day of venlafaxine during eight weeks in severely depressed patients (MADRS>30), escitalopram led to a significantly enhanced improvement (p<0.05). A grouped analysis of two similarly designed studies showed that the difference in favour of escitalopram increased at the same time as the initial severity increased. An analysis of two studies comparing 10 to 20mg per day of escitalopram to 60mg per day of duloxetine in severely depressed patients, revealed the superior efficacy of escitalopram in the sub-sample of severely depressed patients (p<0.01), with significant superiority on each of the 10 items of the MADRS taken singly. Despite the limits of regrouped analyses, all these results underline the fact that escitalopram is at least as effective as the comparators, and notably compared to the two SRINA studied, in the severe forms of depression.