The BCA, the 125-C1qBA, the mRF inhibition assay, Raji cell assay, and the SBA, all assays for ICs, were evaluated for their performance characteristics on the same specially prepared samples (267 sera), and results were compared by reference to a common standard. Differences in reproducibility between the tests were relatively consistent regardless of the parameter of variation examined or the internal standard used to compute results (overall rank order of variation: BCA and SBA less than C1qBA and Raji less than mRF). Larger ICs were overestimated and smaller ICs were underestimated by the assays. Intermediate (11s to 19s) complexes were detected by all the assays, but there were marked differences in sensitivity for this material (SBA greater than mRF greater than BCA greater than Raji greater than C1qBA). The correlation of IC assay results with IgG levels in clinical specimens could not be reproduced in vitro with monomeric IgG, and thus an artifactual influence of IgG levels on IC assay readouts was not the primary reason for this correlation. All assays were influenced by the presence of IgM rheumatoid factors. The effect of other serum manipulations such as polyanions, anticoagulants, and variation in complement could have been predicted from the receptor principle of each assay.