Schizophrenia is a complex mental disorder with unknown aetiology. Both candidate gene and genome-wide association (GWA) studies suggest that the human leukocyte antigen (HLA) system may play a part in development of the illness, but the causal HLA variant(s) remain(s) unclear. Previous studies showed that the DRB1*0101 and DRB1*13 alleles might be associated with a high risk of schizophrenia. Therefore, the present study was undertaken to test their association with the disease by genotyping seven DRB1-tagging single nucleotide polymorphisms (SNPs) in a British population. The results showed that, of the previously reported variants that were associated with schizophrenia, the DRB1*1303 allele was the only one marginally associated with a protective effect on the illness in our sample set (¿ (2)¿=¿4.138, P¿=¿0.042, odds ratio (OR)¿=¿0.42, 95 % confidence interval (CI) 0.27-0.66). Interestingly, a significant association was found for rs424232 (¿ (2)¿=¿9.404, P¿=¿0.002, OR¿=¿0.69, 95 % CI 0.54-0.88), which is a tag SNP for the DRB1*1303 allele and located near to the NOTCH4 gene that is a schizophrenia susceptibility locus confirmed by GWA studies. Analysis with the Haploview program demonstrated that rs424232 was in complete linkage disequilibrium with rs3130297 and rs3131296 present in the NOTCH4 locus. While we have failed to confirm association of the candidate alleles in the DRB1 gene with a high risk of schizophrenia, the present work suggests that the association signal detected in the HLA class II locus may extend a relatively long distance, and more work is needed in order to identify the true causal variants within this region or nearby.