Stings from bees and wasps can cause systemic reactions which can be fatal in some individuals. In these venom-sensitive patients, specific IgE to the venom is produced and is considered to participate in the adverse reactions. This immune response requires antigen presentation by human leucocyte antigens (HLA) class II molecules, which includes DR and DP, which are present on antigen presenting cells. To test for associations between HLA class II DRB1 and DPB1 alleles and life-threatening sensitivity to both bee and wasp venoms. To establish further whether any associations are independent of the atopy phenotype. A total of 33 bee- and 44 wasp-venom-sensitive patients was studied. DRB1 genotypes were determined by single stranded oligonucleotide (SSO) probing of PCR products, and DPB1 genotypes by amplified fragment length polymorphism (AFLP) analysis. Total and specific IgE were measured using the Pharmacia Immunocap, FEIA. Patients with specific IgE to the venom antigens only were termed monosensitive and those with additional specific IgE to HDM and/or GP were termed polysensitive. Allele frequencies were compared to an unrelated control population. The 33 bee-sensitive patients had a greater prevalence of DRB1*07 alleles than the control subjects, 26% vs 14%, with an odds ratio (OR) of 2.1 (95% CI, 1.2-3.7, P = 0.015, corrected for multiple comparisons, Pc = ns). This association was confined to the 15 monosensitive bee patients, who had a 43% DRB1*07 allele frequency when compared with 11% in the 18 polysensitive bee patients, OR 6.1 (95% CI, 1.73-22, P = 0.004, Pc = 0.05), and when compared with a control group of non-venom subjects, 43% vs 16%, OR 3.9 (95% CI, 1.72-9.0, P = 0.002, Pc = 0.02). The 44 wasp-sensitive patients had an increase in the DRB1*11 allele when compared with the control subjects, 13% vs 6%, with an OR 2.2 (95% CI, 1.0-4.6, P = 0.04, Pc = NS), and a decreased prevalence of DRB1*04 alleles, 10% vs 19%, with an OR 0.33 (95% CI, 0.24-0.99, P = 0.04, Pc = NS), but these were not significant when multiple comparisons were taken into account. The DPB1 alleles were not significantly different between the venom sensitive patients and the controls. Patients monosensitive to bee venom had a significantly greater prevalence of DRB1*07 alleles than the non-venom, control population suggesting that IgE responses in these patients may, in part be controlled by immune response HLA class II genes. These results are also suggestive of wasp-sensitive patients having a higher prevalence of DRB1*11 and a lower prevalence of DRB1*04 than the control population.