Overexpression of cathepsin D in several types of carcinoma in women appears to be associated with a poor clinical course. In this prospective investigation, cathepsin D levels in 170 specimens of normal and neoplastic human tissues were determined simultaneously by enzyme immunoassay (EIA) and immunoradiometric assay (IRMA) to allow comparisons in multicentric studies, such as cooperative clinical trials. Nonmalignant uteri and specially prepared reference powders were also evaluated. Linear regression analysis between the two assays for all specimens [EIA = 0.87(IRMA)-3.18] demonstrated a correlation coefficient (r) of 0.99 (P < 0.001). When malignancies were categorized by the tissue origin (i.e., breast, uterus, ovary, lymph node, and colon), highly significant correlations were also observed (regressions slopes ranged from 0.58 to 1.02). Intra- and interassay controls conducted for the new EIA procedure gave CV% ranging from 4.4 to 10.2, which was similar to the IRMA test for cathepsin D. The results of both assays correlated well and were highly reproducible. Either assay may be used with confidence that comparable cathepsin D values will be obtained in a wide range of tissue biopsies.