Because of the inherent difficulties in allele assignment with HLA-DR serological typing, in 1993 our organ procurement organization-based HLA laboratory replaced serology with the molecular method of polymerase chain reaction using sequence-specific primer mixes (PCR-SSP) to type for DR and DQ at a resolution level equivalent to that of serologically defined antigens. In this study, we compared the incidence of DR blanks, where allocative homozygosity occurred, and graft outcome during our serology epoch (1987-1993) with that of our molecular epoch (1993-1996). The incidence of DR blanks by PCR-SSP (17.0%; 138/1101) was significantly lower (P<0.005) than in the serology epoch (21.5%; 569/2647). Although DQ is not a component of the allocation algorithm, the incidence of blanks in the molecular era (21.9%; 196/895) was 46% lower (P<0.001) than in the serology epoch (40.8%; 931/2277). Graft survival in 163 cadaveric renal transplant recipients for whom molecular DR allocation occurred (patient and donor were molecularly typed) showed that PCR-SSP typing had no significant effect on 2.5-year graft survival for patients mismatched for 0 (97%), 1 (90%), or 2 (94%) HLA-DR antigens (P=0.4; log-rank). In conclusion, molecular typing lowered the rate of DR and DQ blanks, but molecular matching for HLA DR and DQ did not influence graft outcome at 2.5 years.