A sensitive catalytic assay for placental alkaline phosphatase (PLAP) was used to quantify enzyme levels in sera from women with malignant gynecological tumors, i.e. cervical carcinoma, ovarian carcinoma and carcinoma of the breast. Values were compared with those of a reference group of 155 healthy individuals. In this group, basic mean levels of the enzyme amounted to 0.06 +/- 0.08 mumoles X min-1 X 1(-1) (mean +/- SD), which corresponds to 0.3 +/- 0.4 ng/ml (mean +/- SD) purified placental alkaline phosphatase. Enzyme levels were not significantly affected by age, whereas sex gave slightly higher values for women in all age groups. This difference was statistically significant (p less than 0.001). The discriminating capacity of PLAP as a tumor marker was analysed with this assay and compared in sensitivity with a radio-immunoassay. A cut-off level of 0.2 mumoles X min-1 X 1(-1), corresponding to 1 ng/ml enzyme protein as measured by the radio-immunoassay, detected 23-68% of the tumor patients, giving false-positive results from the control group in the order of 5%. Corresponding values at a cut-off level of 0.3 mumoles X min-1 X 1(-1) (1.5 ng/ml) gave values 20-44% and 2%, respectively. The present investigation emphasizes that sensitivity in assays of placental alkaline phosphatase, whether catalytic or immunological, must not exceed ng-level.