For 39 controlled studies on the analgesic effect of antidepressants, a meta-analysis was conducted to get an estimation of the effect size, and to obtain a sight on the possible modes of action and the methodology used. The mean size of the analgesic effect was 0.64. It means that the average chronic pain patient who received an antidepressant treatment had less pain than 74% of the chronic pain patients who received a placebo. This quantification, however, is only as good as the studies on which it is based, and it could be differentiated for each of the pain syndromes and antidepressants examined. Real analgesic qualities of antidepressive agents seemed to offer the most plausible and economical explanation for the effect, but the predominant importance of serotonin reuptake blocking was not confirmed. Finally, the meta-analysis appeared to be fruitful for the generation of new hypotheses, for making some recommendations for future research, and for proposing some provisional guidelines for the clinical use of antidepressants in chronic non-malignant pain.