Bipolar affective disorder (BPAD) is a complex psychiatric disorder with a major genetic contribution. Abnormalities in serotonergic function have been implicated in its aetiology. The 5HT2A receptor (5HT2AR) gene is a strong candidate gene for involvement in BPAD, but previous association studies have reported conflicting results. These data are difficult to interpret because most negative results were obtained with small samples. The aim of this study was to test the association between the 5HT2AR gene and BPAD in a large West European sample. We studied the −1438G/A and the His452Tyr polymorphisms, for haplotype analysis to increase both informativity and the likelihood of detecting an association between BPAD and the 5HT2AR gene. We analysed the genotype, allele and haplotype distributions of two 5HT2AR gene variants in a population of 356 BPAD patients, which we compared with 208 healthy controls. We also carried out exploratory analysis in clinical subgroups of patients defined according to personal history of mood disorders, suicidal behaviour, comorbid psychiatric disorders and family history of affective disorders. We found no difference between BPAD patients and controls for allele, genotype and haplotype distributions. Exploratory analysis in subgroups of BPAD patients showed only a marginal difference in haplotype distribution between controls and BPAD patients with antidepressant-induced mania (P = 0.018). This difference was not significant after correction for multiple testing. Our study suggests that the 5HT2AR gene is unlikely to be involved in genetic susceptibility to BPAD but should be further investigated in a pharmacogenetic study. © 2004 Wiley-Liss, Inc.