A previously reported direct enzymatic method for cholesterol [ Clin. Chem. 20, 470 (1974)] was critically evaluated in the clinical laboratory. Other methods compared to this enzymatic procedure were: (a) the manual Abell—Kendall procedure, giving a regression slope (proportional sensitivity) of 0.998 and a negative bias of 9.7 mg/dl in the enzymatic procedure; (b) Lipid Research Clinic procedure, giving a regression slope of 1.005 and a negative bias of 8.8 mg/dl in the enzymatic procedure; and (c) the SMA 12/60 (Technicon) method, giving a regression slope of 0.881 and a negative bias of 20.8 in the enzymatic procedure. Replicate analyses of pooled serum by the enzymatic procedure demonstrated the following inter-run precision: mean = 178 mg/dl, SD = 4 mg/dl, CV = 2.2%. Bilirubin caused a slight positive interference (1 mg of bilirubin per deciliter appears as 1-2 mg of cholesterol). Hemolysis, lipemia, and a series of steroid drugs caused no interference. A group of 465 normal subjects was tested by the enzymatic procedure. This population had a mean cholesterol of 203 mg/dl. The 2.5 percentile was 130 mg/dl, and the 97.5 percentile was 316 mg/dl.