Concordance to antidepressant response in members of the same family is a common observation in clinical practice. However, few published data support this view; moreover families with affected members responder to the same antidepressant have been poorly studied. We have analyzed 45 pairs consisting of one mood disorder fluvoxamine double-responder proband and one first-degree relative with known outcome to fluvoxamine treatment. Among 45 pairs 30 (67%) were concordant for good response to fluvoxamine. In family pedigrees of concordant pairs we found a significantly higher distribution of bipolar forms in secondary cases than in families of non concordant pairs (14.9% vs 3.9% P = 0.039) suggesting that concordance to antidepressant therapy could select families with higher genetic loading.