Data presented by Straus and Bacon (Quart. Jour. Stud. Alcohol 12: 231-60, 1951) have been subjected to statistical analysis. Contrary to their conclusion, the clinics represented in their study are not generally homogeneous with regard to the marital status of their patients. Four possible areas of difference between clinics are suggested: (1) intake policies, (2) regional differences in marital status, (3) recording practices, (4) sources of patient referral. In studies of alcoholics the inadequate analysis of data may serve to hide real differences and make future research more difficult.