Summary Fifteen human and two rabbit sera that contained anti-bovine serum albumin (BSA) as measured by the ammonium sulfate test were studied with a variety of other methods for detecting antibody. These included radioimmunoelectrophoresis, P-80 precipitin test, microimmunodiffusion, hemagglutination, passive cutaneous anaphylaxis and the Prausnitz-Küstner passive transfer for reaginic activity. The rabbit anti-BSA sera bound amounts of I131-BSA comparable to the human antisera and gave strongly positive reactions when tested with each procedure. Radioimmunoelectrophoresis studies confirmed that each of the human sera selected on the basis of the ammonium sulfate test contained antibodies capable of binding I131-BSA, but among all the other techniques there was considerable variation with regard to their capacity to detect and measure anti-BSA activity in the human sera. The P-80 test revealed that the human sera had reduced precipitating efficiency as compared to similar rabbit anti-BSA sera. Microimmuno-diffusion was relatively insensitive in that it detected antibody in only one human serum. The hemagglutination test was positive in 10 of the 15 sera, but the titers were relatively low and did not reflect the major differences observed with the ammonium sulfate test. Reaginic activity was present in four human sera and only three of these contained γA-globulin which bound I131-BSA when tested with radioimmunoelectrophoresis. Conversely, one human anti-BSA serum contained γA-globulin which bound I131-BSA but did not have reaginic activity. Thus, reaginic activity was not necessarily related to the presence or absence of γA-globulins with the capacity to bind I131-BSA. The passive cutaneous anaphylaxis tests were uniformly negative employing the human anti-BSA sera. Of the techniques used in this study, only radioimmunoelectrophoresis and the ammonium sulfate method detect the primary interaction between antigen and antibody. In contrast, the other in vitro and in vivo tests depend upon measuring secondary events which in varying degree may occur sometime after the primary antigenantibody reaction.