URD and patients matched by HLA-A and -B serology and -DRBI are known to have disparities for HLA-DQBI alleles. We studied 129 patients and 431 URD to examine whether DNA typing for HLA-A, -B, and -DRBI has any influence on the degree of HLA-DQBI matching. The DNA typing was performed by PCR-SSOP method. Of 431 URD, 102, 169, and 160 were 6/6, 5/6, and 4/6 DNA matches, respectively. Of 6/6 URD. 86 (84.3%) were matched for both and the rest (15.7%) were mismatched for one HLA-DQBI allele. In the 5/6 group, 115 (68.0%), 51 (30.2%), and 3 (1.8%) and in the 4/6 group, 86 (53.8%), 61 (38.1%), and 13 (8.1%) URD were matched for both, one, and none of HLA-DQBI alleles, respectively. Compared to the 6/6 group, a higher degree of HLA-DQBI disparity was seen in the 5/6 (pO.OI) and 4/6 (p<0.01) groups. The degree of disparity in the 4/6 was higher than in 5/6 (p<0.01). Next we examined the effect of matching at individual loci. The URD (n=431) were stratified for HLA-A, -B, and -DRBI matching as follows: MA-matched for both HLA-A alleles (n=296); MMA-mismatched for HLA-A alleles (n"135); MB-matched for both HLA-B alleles (n=270); MMBmismatched for HLA-B alleles (n=16l); MD-matched for both HLA-DRBI alleles (n275); and MMD-mismatched for HLA-DRBI alleles (n=56). Within MA group, 194 (65.6%), 88 (29.7%), and 14 (4.7%) and in MMA group, 93 (68.9%), 40 (29.6%). and 2 (1.5%) URD were matched for both, one, and none of HLA-DQBI alleles. In MB group, 180 (66.7%), 77 (28.5%), and 13 (4.8%) and in MMB group, 107 (66.4%), 51 (31.7%), and 3 (1.9%) URD were matched for both, one, and none of HLA-DQBI alleles, respectively. The degree of HLA-DQBI disparity was similar between MA and MMA (p=0.26) as well as between MB and MMB (p-0.27) groups. In MD group, 215 (78.2%), 60 (21.8%), and zero URD and in the MMD group, 72 (46.1%), 68 (43.6%). and 16 (10:3%) URD were matched for both, one, and none of HLA-DQBI alleles. The odds ratio of HLA-DQBI matching for MD vs. MMD group was 4.15. The degree of HLA-DQBI disparity was significantly higher in MMD compared to MD group (p<0.01). We conclude that URD matching for HLA-A, -B, and -DRBI, even by DNA typing, does not guarantee HLA-DQBI matching. Furthermore. DNA matching for HLA-A and -B has no influence on the degree of HLA-DQBI matching. Conversely, the level of HLA-DRBI matching positively influences the degree of HLA-DQBI matching reflecting linkage disequilibrium between the two loci. The outcome studies will determine the clinical importance of the above observations.