We evaluated the analytical performance of a new monoclonal immunoradiometric assay ("M-PSA") for prostate-specific antigen ("Tandem"; Hybritech Inc.) in comparison with a monoclonal immunoradiometric assay ("M-PAP") for mass measurement of prostatic acid phosphatase ("Tandem") and with a conventional enzyme-activity assay ("E-PAP") for prostatic acid phosphatase (EC 3.1.3.2). For M-PSA, the CVs were 1.3-3.0% within-run and 3.0-4.9% between-run. The minimum detectable mass concentration was 0.10 microgram/L, and linearity extended to 100 micrograms/L. The reference interval for M-PSA in 178 healthy men was 0-2.8 micrograms/L. Serum specimens from men with prostatic disease (primarily prostatic carcinoma and benign prostatic hypertrophy) were assayed by the three methods. Correlation was best between mass measurement (M-PAP) and enzyme activity (E-PAP) for prostatic acid phosphatase (r = 0.958). Results for PSA did not correlate well with those for either M-PAP (r = 0.629) or E-PAP (r = 0.387). PSA was increased in a higher percentage of specimens from men with earlier (clinical stage B) prostatic carcinoma than were results from either assay for PAP.