BACKGROUND: In common with other animal models of psychopathology, the forced-swimming test (FST) suffers from the fact that it involves normal animals. Moreover, powerful antidepressant drugs such as the selective serotonin reuptake inhibitors have been found to give false negatives in this behavioral test. METHODS: To circumvent these theoretical and practical difficulties, we studied the interindividual variability of the behavioral reactivity of rats in the FST. The effects of fluoxetine treatment or of a stressful experience (repeated testing in the FST) were analyzed on various behavioral responses. RESULTS: The following observations were made in replicated experiments: 1) a dimensional behavioral response from passivity to high reactivity in the FST; 2) an antidepressant-like effect of fluoxetine only in a subgroup of animals categorized as low responders on the dimension of passivity-reactivity; and 3) a switch toward passive responses following a past experience of stress, which was corrected by fluoxetine treatment. CONCLUSIONS: It is concluded that a dimensional approach could improve the screening of antidepressant drugs and could aid the development of new ones by identifying the biobehavioral characteristics of responder and nonresponder subjects.