The short variant of the serotonin transporter gene (SERTPR) has been consistently associated with a poorer response to treatment with various selective serotonin reuptake inhibitors (SSRIs). Antidepressant response is not a unitary phenomenon, however, and we here hypothesized that the SERTPR effect could be specific to some types of symptomatology. The sample comprised 281 inpatients affected by mood disorders and treated for major depression with SSRIs. The total depressive scores for all patients were analyzed in previous reports, but symptomatologic clusters were not examined previously. The 21-item Hamilton Rating Scale for Depression (HAM-D) was administered to evaluate depressive symptoms at baseline and weekly over 6 weeks of treatment. All patients were genotyped for the SERTPR polymorphism. Compared with patients with the SERTPR l/l and l/s polymorphisms, s/s patients showed a selective and slower improvement of depressive "core" and somatic anxiety symptoms, but they did not differ from other patients regarding other symptomatologic clusters such as insomnia and motor retardation. These findings support the view that response to SSRIs is not a unitary phenomenon and that improvement of symptomatologic clusters as, at least in part, genetically driven. SERTPR may be hypothesized as concurrently participating to the activity of anatomic brain regions differentially involved in depression and somatic symptoms of anxiety; however, further studies are required to examine these complex interactions.