Determinations of SACE activity were performed in 80 patients with sarcoidosis, 55 normal controls, and 29 patients with asthma, by the spectrophotometric method of Cushman and Cheung. SACE levels were significantly higher in both untreated and steroid-treated patients with sarcoidosis than in normal controls: 46.2 +/- 20.6 (S.D.), nm/min/ml, 38.1 +/- 23.1, and 26.8 +/- 1.8, respectively. There were no significant differences between steroid-treated and untreated patient groups. However, an inverse correlation was observed between SACE levels and steroid dose, suggesting the possibility of a dose dependency for steroid-induced depression of SACE. Elevated SACE levels (2 S.D. above mean controls) were present in 67% of untreated sarcoidosis patients. The prevalence of elevated SACE levels was not significantly higher when the patient population was examined with respect to duration of disease, radiological stage, and the presence of abnormality in pulmonary function tests. When patients were divided according to the frequency of clinical criteria of disease activity, the presence of two or more criteria was associated with elevated SACE levels in 88% of patients. But SACE levels were elevated in 32% of patients judged to have dormant disease by clinical criteria. SACE levels had an 81% accuracy in prediction of disease activity and a 79% accruacy for prediction of inactivity. It was concluded that SACE elevations have definite diagnostic value in sarcoidosis and are helpful in establishing the presence of disease activity but are not sufficient to fully separate active from inactive disease groups.