We have recently shown that an as yet unidentified gene within or in the vicinity of the HLA complex, in linkage disequilibrium with microsatellite D6S2223, modifies the risk to develop type 1 diabetes independently of HLA-DR and -DQ genes. This microsatellite is located 2.5 Mb telomeric to HLA-F and particular alleles at this microsatellite modifies the risk encoded by the high-risk DRB1*03-DQA1*0501-DQB1*0201 (hereafter called DR3) haplotype. Coeliac disease and type 1 diabetes share some susceptibility HLA class II haplotypes, in Scandinavia particularly the DR3 haplotype. We therefore investigated whether the marker D6S2223 might also be associated with coeliac disease. In order to keep the contributions from the DRB1-DQA1-DQB1 genes constant (i.e., eliminate the effects of linkage disequilibrium to disease associated DR and/or DQ alleles), we only used cases and controls being homozygous for DR3. We found the frequency of allele 3 at D6S2223 to be reduced among patients with coeliac disease compared to controls, to a similar extent as seen in type 1 diabetes, which could not be explained by a different distribution of HLA-B alleles (as ascertained by typing for the MIB microsatellite). This negatively associated allele 3 at D6S2223 occurred in a homozygous combination at a significantly lower frequency among patients than controls. Thus, allele 3 at D6S2223 on DR3 haplotypes is associated with reduced susceptibility for development of both type 1 diabetes and coeliac disease. This suggests that a gene(s) in the vicinity of D6S2223 is involved in the pathogenesis of both of these immune-mediated diseases.