The new CA 125 II (Centocor) serum assay utilizing the M11 mouse monoclonal antibody as capture antibody and OC125 as tracer antibody, was investigated for its technical and clinical performance against the original CA 125 assay. The CA 125 II test revealed a quadruple increase in signal-to-noise ratio, good intra- and interassay precision (with CVs < 5% and 7%, respectively), improved dilution linearity, and a minimal detectable dose of 0.38 units/mL. Sera were obtained from healthy females (n = 192), women with benign conditions (n = 208), and patients with various cancers (n = 379). Both assays measured highly similar CA 125 distributions with equal reference ranges and nearly identical positivity (> 35 units/mL) rates, resulting in similar receiver-operating characteristic curves and monitoring graphs. Linear regression analysis of results by the two assays (CA 125 = x, CA 125 II = y) in ovarian cancer patients showed, for CA 125 assay values between 0 and 1000 units/mL, a slope of 1.00 and a y-intercept of 12.6 (n = 254, r = 0.8617, P < 0.0001). The heterologous CA 125 II assay appeared to be more accurate in patients who had human anti-mouse antibodies after immunoscintigraphy. The CA 125 II immunoradiometric assay is sensitive and reliable for measuring serum CA 125, and fully retains the cutoff values of 35 and 65 units/mL that were defined with the original CA 125 immunoradiometric assay.