We report the evaluation of a new commercial kit for the determination of pancreatic lipase activity. The kit is based on the use of a 1,2-diglyceride as substrate and a specific monoglyceride lipase. The detection step is the continuous colorimetric measurement of hydrogen peroxide produced from glycerol by glycerol kinase, glycerol-3-phosphate oxidase, and peroxidase reactions. The procedure appears to be precise (between-day CV < 9%) and the results show good correlation with those obtained by alternative procedures (vs turbidimetry, r = 0.965; vs ultraviolet absorbance-enzymatic method, r = 0.995; vs Ektachem, r = 0.976; vs immunometry, r = 0.970). However, the method is susceptible to interference by increased concentrations (> 4.5 mmol/L) of serum triglycerides. We estimated the reference interval for healthy adults to be 8-44 U/L. When we evaluated clinical efficacy by using receiver-operating characteristic curves and the overlap index, no significant differences were found between the commercial kit and a common turbidimetric assay for diagnosing patients with acute pancreatitis; both methods performed satisfactorily.