The objective of this study was to evaluate the diagnostic utility of the measurement of the serum activity of angiotensin-converting enzyme (SACE) as a cancer marker. This case-control study included 135 patients with cancer of different sites, confirmed histologically, and 145 controls (107 normal individuals plus 38 chronically ill patients with nontumoral diseases). Determination of SACE activity was done by a spectrophotometric method using as substrate the synthetic tripeptide N-(3-[2-furyl]acryloyl)-L-phenylalanylglycine. There were no sex- or age-related variations in SACE activity. Mean SACE activity in 107 normal controls was 51.6 U/l (95% C.I., 50.1-53.1); in 145 nontumoral individuals, including 38 chronic nonmalignant diseases plus 107 normal controls, 51.5 (50.1-53.1) and in malignant tumors 35.7 (32.8-38.5). There was no statistically significant difference between chronic diseases and normal controls (p > 0.05); but there was one between cancer patients and nontumoral individuals, normals and chronic nontumoral diseases. The mean of SACE activity values by tumoral site are (U/l; 95% C.I.): breast, 41.3 (36.2-46.5); gastrointestinal 31.5 (24.3-38.8); head and neck, 32.3 (26.7-37.8), and lung 27.6 (21.6-33.6) (p < 0.001). The means by clinical stage are: complete remission, 58.0 (53.7-62.3), significantly higher than in normal controls (p < 0.001); local disease, 40.56 (34.5-46.5); locoregional disease, 35.09 (30.7-39.4); metastatic disease, 23.04 (19.5-26.5), and in relapse at diverse stages, 30.86 (25.1-36.5). In clinical active cases, there is a statistically significant decrease of SACE activity, especially in metastatic disease (p < 0.001). The calculated cutoff value, excluding complete remission cases, is 40.7 U/l, with sensitivity of 69.5% and specificity of 91.6%. We conclude that there is a decrease of SACE activity in cases of clinically active cancer and an increase in clinical complete remission.