The DQA1*0501 and DQB1*0201 alleles (hereafter DQ2) confer genetic susceptibility to celiac disease (CD). Some studies have indicated that the DPB1, DMB, and TAP loci, that are located close to the DQ genes, could also together with DQ or independently confer genetic susceptibility to CD. Some others have claimed that these associations result merely from linkage disequilibrium, a hallmark of the MHC, that often makes the precise mapping of susceptibility genes difficult. To evaluate further the role of class II genes in CD, we analyzed segregation of DPB1 alleles in families with CD. In particular, we analyzed families where one of the parents was homozygous for the DQ2 risk allele but heterozygous for DPB1*0101, an allele claimed to be an additional risk allele. We reasoned that if DPB1*0101 would not play a role in CD, then patients should inherit the DQ2 haplotypes randomly from a homozygous parent. We here present evidence that, in all 6 informative families, those DQ2 positive haplotypes that also include the DPB1*0101 allele, rather than those without DPB1*0101, are predominantly segregated to the index patient from the parent homozygous for the DQ2 risk marker (p = 0.03 as compared to their healthy siblings). If confirmed in larger studies the results indicate that despite DQ2 other genes in or near the MHC may associate with CD.