Seasonal variation in the births of patients with schizophrenia is a consistently replicated epidemiological finding. Few studies have investigated this phenomenon among patients with a diagnosis of affective disorder. The majority of season of birth studies have employed the chi square test for statistical analysis, a method that has been subject to some criticism. Using a Kolgomorov-Smirnov type statistic, the quarterly birth distribution of 6,646 patients with an ICD 9/10 diagnosis of affective disorder were compared to the general population. Only the births of those individuals with unipolar forms of affective disorder (n = 4,393) differed significantly from the general population, with significant excesses and deficits in the second quarter and fourth quarter respectively. These results were not altered by application of the displacement test. © 1998 Elsevier, Paris