The standard lactose tolerance test involves measuring a patient's blood glucose after the ingestion of lactose. If the patient has lactase deficiency and is unable to hydrolyze lactose and absorb its monosaccharides, glucose and galactose, the blood glucose does not usually increase greater than 20 mg/100 ml. Since factors other than the absorption of glucose can cause an increase in the blood glucose of greater than 20 mg/100 ml in a diabetic, this test could be unreliable when it is performed on a diabetic. The present study was performed to determine whether the lactose-ethanol tolerance test could be used to diagnose lactase deficiency in diabetics. This test involves measuring the blood galactose level, instead of the blood glucose, and the administration of ethyl alcohol to a subject prior to the test to delay the clearance of galactose from the circulation. The results indicate that the standard lactose tolerance test in which the blood glucose is measured is unreliable when performed on insulin-dependent diabetics, but that it can be reliable when performed on non-insulin-dependent diabetics. The lactose-ethanol tolerance test gave results in each type of diabetic which were qualitatively similar to those of nondiabetics. It was concluded that the latter test is a useful screening test for lactase deficiency in diabetics.