HLA-specific antibody, present before or after transplantation, may adversely effect graft outcome. Antibody testing by cytotoxicity (CYT) is laborious, requires viable lymphocytes, does not differentiate non-HLA cytotoxic antibody, and cannot be used readily on specimens from patients being treated with cytotoxic antibodies. We have evaluated PRA-STAT, an antibody screening kit that uses an ELISA test with soluble HLA class I molecules as targets. We performed 219 tests on a variety of serum specimens, 128 of which were also tested by CYT. There was a highly significant correlation (r = 0.78, P < 0.001) between PRA-STAT (PS) and CYT for the detection of IgG antibodies. Of 66 sera reactive in both assays, 18% had identical specificities defined in both, 27% were more reactive in PS than in CYT, 8% were more reactive in CYT, and 47% had different specificities in the 2 assays, with overlap in slightly more than half the cases. Of 13 sera reactive only in PS, 2 were from non-transfused, nontransplanted males with no evidence of lymphocyte-reactive antibody by antiglobulin tests. PS uses an IgG-specific conjugate, therefore IgM class I-specific antibodies cannot be identified--however, their presence does affect test outcome. This, as well as the panel composition and interlot reproducibility, are areas we believe need to be addressed. The PRA-STAT system is rapid, does not require viable cells or complement, and can be automated in part. Resolution of the problems identified here and availability of an IgM-specific conjugate should make this test system a valuable tool in histocompatibility testing.