Although the excess of schizophrenic births in the winter and early spring has been replicated and some non-conclusive work supports the same seasonal birth trend in patients with major affective disorders, the aetiopathogenetic foundations of this phenomenon remain uncertain. The primary role of perinatal seasonal factors that predispose to the development of schizophrenia via induction of brain damage has been invoked, as has a tendency for patients to conceive during the spring and early summer. In order to test these two hypotheses, cerebral ventricular size and cortical atrophy in 206 schizophrenics and 107 patients with major affective disorders were assessed by CT and analysed in relation to month of birth. Compared with schizophrenics born during the remainder of the year, those born between December and April, particularly in cases lacking a family history of schizophrenia, showed increased chances for ventricular enlargement, but not for cortical atrophy. No association between season of birth and central or cortical atrophy was found for patients with major affective disorders. This suggests that the brain-damaging effect played by perinatal seasonal factors has both a disease and an anatomical specificity.