Angiotensin-converting enzyme (ACE) activity in serum is used as an aid to the diagnosis and follow-up of patients with sarcoidosis. A theoretical limitation of measurements of activity is that these may be affected by the presence of pharmacologic or endogenous inhibitors of ACE. Immunoassays of ACE concentration avoid this problem and, when combined with tests of ACE activity, permit calculation of specific activity of ACE. In this study, we set out to develop a sensitive radioimmunoassay for ACE to compare results obtained with this method with results of ACE activity and calculated ACE specific activity in patients suffering from a variety of lung diseases. In a group of control subjects (n = 32), the ACE concentration was 453.7 +/- 159.8 (SD) ng/mL; 95% confidence interval (CI), 398.34 to 509.06, but levels were significantly elevated in sarcoidosis (979.3 +/- 558.6 ng/mL; 95% CI, 827.5 to 1,131.1; n = 51; p < 0.001 vs control subjects), silicosis (646.5 +/- 239.1 ng/mL; 95% CI, 544.2 to 748.8; n = 21; p < 0.01), and miliary tuberculosis (647.0 +/- 217.1 ng/mL; 95% CI, 551.9 to 742.1; n = 29; p < 0.01). The levels were normal in COPD, interstitial pulmonary fibrosis, and active cavitary pulmonary tuberculosis. The overall correlation between ACE activity and concentration measurements was strong (r = 0.93). No evidence of endogenous ACE inhibition was observed in any of the disease categories studied except in COPD where an elevation of ACE specific activity was observed, raising the possibility that in this condition different isozymes of ACE with higher specific activity might be released.