To determine whether the lipase:amylase ratio differentiates alcoholic from nonalcoholic pancreatitis, we conducted a retrospective review of charts with the diagnosis of acute pancreatitis at the George Washington University Medical Center between January 1988 and July 1990. A total of 446 charts were reviewed. For a patient to be included in the subsequent analysis, the following criteria were met: 1) the patient had typical symptoms of pancreatitis, 2) serum amylase and lipase were analyzed on admission, and 3) a computerized tomographic (CT) scan or ultrasound of the abdomen was obtained within 72 h of admission. Forty-seven charts satisfied the requirements for inclusion in the study. Data collected from the charts included history of alcohol consumption, age, sex, race, admission serum amylase and serum lipase (from this the amylase:lipase ratio was calculated), peak serum amylase and serum lipase, and number of days of abdominal pain before admission. Patients with alcoholic pancreatitis had significantly lower serum amylase levels and significantly higher lipase:amylase ratios than those with nonalcoholic pancreatitis (p < 0.01). There was no difference in the serum lipase between the groups. The higher the lipase:amylase ratio, the greater the specificity of alcohol as the etiology of acute pancreatitis. Only patients with alcoholic acute pancreatitis had lipase:amylase ratios > 5.0 (sensitivity 31%, specificity 100%). Our data point to the clinical utility of the lipase:amylase ratio in differentiating alcoholic from nonalcoholic acute pancreatitis. Prospective studies will be needed to confirm the clinical utility of this ratio.