The influence of human anti-OC125 antibodies formed after multiple infusions of OC125 F(ab'), fragments on the apparent levels of CA 125 measured with four different tests were examined in two ovarian cancer patients. With the homologous Assay 1, involving only OC125 antibodies, false increases of CA 125 values were observed after infusion of OC125 fragments, which completely covered real CA 125. In contrast, with Assay 2, 3 and 4, which involve no OC125 antibodies as capture antibodies, only slight false increases occurred in the presence of very high anti-OC125 antibody concentrations. Interference was eliminated by addition of non-specific murine IgG in Assay 2 and 4, but not in Assay 1 and 3 indicating that the false increases in Assay 1 and 3 were caused by anti-idiotypic anti-OC125 antibodies. In the presence of elevated real CA 125, with Assay 2 and to a considerably lesser degree with Assay 4, an inhibitory effect of anti-OC125 antibodies became evident leading to false decreases of CA 125 values. In Assay 4 reduction of assay response was eliminated by addition of nonspecific murine IgG. The results confirm that all available CA 125 tests are influenced by interference with human anti-OC125 antibodies. Thus, CA 125 levels in patients who have been treated with OC125 fragments should be interpreted with care.