Genetic and environmental factors, as well as their interactions, are likely to be involved in psychiatric disorders. Considerable progress has been made in association and linkage studies with various candidate genes, at times with conflicting or ambiguous results. An environmental factor that has persistently shown associations with several psychiatric and neurological disorders is the season of birth. If it is the interaction of a specific gene allele with a specific season of birth that constitutes an increased (or decreased) risk for a disorder, then the individuals with this disorder are likely to have a season of birth variation in this gene allele. We investigated the variations in TPH, 5-HTTLPR and DRD4 gene polymorphisms according to seasonality of birth in 954 patients with unipolar affective disorder, bipolar affective disorder, and schizophrenia, respectively, and in 395 controls. We first analyzed season of birth variations in the gene alleles with one cycle or two cycles per year, and then compared specified birth seasons with each other. We found season of birth variations in these gene alleles that were different for different psychiatric disorders. Significant differences between cases and controls could be obtained when restricting the analysis within certain birth seasons but not within others. Our results thus suggest an interaction between the seasons of birth and the expression of the candidate genes, and that season of birth is a confounding variable when investigating the role of the candidate genes in susceptibility to psychiatric disorders.