Human anti-murine antibody titres following patient exposure to the monoclonal antibody Orthoclone OKT3 (muromonab-CD3) are determined by laboratories using diverse analytical methods which are not standardized and whose concordance is not established. A multicentre study group therefore compared testing for IgG anti-OKT3 antibody among seven laboratories. A set of 270 sera was obtained from 30 heart, 30 kidney and 30 liver transplant recipients with no previous exposure to OKT3 who were receiving OKT3 for induction immunosuppression. Sera were collected from each patient prior to and at 24 +/- 2 days and 31 +/- 2 days following initial OKT3 exposure. Identical aliquots of all 270 sera were tested for IgG anti-OKT3 antibody by each laboratory. In addition, the limit of detection of each laboratory's method was estimated by titration of an affinity-purified IgG anti-OKT3 reference material of known concentration. Anti-OKT3 antibody formation differed greatly among the three organ groups. Cardiac patients demonstrated the least sensitization and almost exclusively lower titres, while kidney recipients had more frequent and higher titre antibody formation. Liver recipients yielded the highest sensitization rate and the most frequent high titre sera. Importantly, the seven laboratories differed widely in the number of pretreatment sera reported as positive (ranging from 0% to 41% among laboratories), the number of post-OKT3 sera reported as positive (17-63%), the number of post-OKT3 samples with titre > or = 1000 (2-31%), and the number of patients sensitized 19-69%). Concordance among laboratories was highly variable, with interlaboratory agreement ranging from 38% to 83% on the sample titres assigned to 180 post-OKT3 sera. Many of the discordant results were consistent with differences in the limit of detection of the analytical methods, which ranged from 0.19 microgram/ml to > or = 15 micrograms/ml, a nearly 100-fold difference among laboratories. This study demonstrated the presence of both good concordance and significant discordance among laboratories in determining human anti-mouse antibody titres, and demonstrated that common titre categories (100, 1000, 10,000) were not equivalent among laboratories. The level of concordance among methods should be considered when comparing anti-OKT3 antibody results from different centres and their correlation with clinical events. Universal comparative testing, patterned after proficiency testing programmes, is needed to assess differences among laboratories and to bring uniformity and a sound interpretative basis to this field of testing.