Anti-class I IgG can be detected by complement-dependent cytotoxicity (CDC) and by ELISA. We compared ELISA and CDC for both class I and class II antibodies on method agreement and relation to rejection-free and graft survival.Peak, current, and posttransplant sera (n=429) of 143 renal allograft patients were tested by National Institutes of Health technique (NIHT), two-color fluorescence (TCF), and ELISA. Method agreement was assessed by intraclass correlation coefficient (ICC). Rejection and graft survival were analyzed by uni- and multivariate techniques. The screening results for each serum were compared, as was the change in result of current to posttransplant serum.The ICC of ELISA and NIHT was insufficient; it was lower for TCF than NIHT. Graft survival was not related to the result of any assay. Rejection-free survival was related to ELISA and NIHT in current and posttransplant serum. With the NIHT, the change in percent panel-reactive antibody (%PRA) correlated better with rejection than it did with ELISA. The combined antibody status of current and posttransplant serum was a risk factor for rejection in all assays, and for TCF also in multivariate analysis. The rejection rate was higher if the posttransplant serum was ELISA-negative/CDC-positive, rather than ELISA-positive/CDC-negative. For ELISA, class I specificities (and not %PRA) in peak and current sera were related to rejection, even if the antibodies were not donor-directed. In the case of the National Institutes of Health technique (NIHT), %PRA and not specificity was related to rejection. Class II antibodies were never related to rejection.ELISA and NEIT are complementary screening techniques in this patient population. They are of equal predictive value for rejection. The optimal strategy in combining these techniques must be determined.