The birthweights of 450 schizophrenic patients born in 1971-1978 were compared with those of individually matched controls from the same birth cohort. Schizophrenics born in the second quarter of the year (April to June) had a significantly lower birthweight than their controls and this was not attributable to differences in the duration of gestation. Analogous proband/control differences were not found in 301 patients with affective psychoses, but there was a similar, much smaller, seasonal fluctuation in birthweight in the general population. These findings may be a clue both to the genesis of the "season of birth effect" and to the identity of the intrauterine influences contributing to the aetiology of schizophrenia. Folate deficiency may be implicated.