Patients (N = 114) consecutively entering a medical service with ulcer, cirrhosis or pancreatitis, currently drinking and not currently active in alcoholism treatment were randomly assigned to motivational intervention (MI) or to a control group. Increased utilization of alcoholism programs and self-reported sobriety at 10 weeks were assessed. MI consisted of three separate discussions of the relationship of the patient's disease to continued drinking and the compassionate offer of treatment. Two persons skilled in treatment also met with each MI subject and discussed treatment possibilities for them, facilitating entrance if desired. Patients in both the MI and control group were treated for their medical condition by a medical team and alcoholism treatment was always recommended. Outcome was evaluated for the period from the 10th to the 16th week after return to the community by interview of patient and household contacts and by the keeping of appointments. There were no differences between the control and MI groups, with at least 38% remaining sober for the 10-week interval; the study size was sufficient to detect reliably a 30% improvement. We conclude that additional motivational intervention to this level was not beneficial to the hospitalized alcoholic with disease. There was a statistically significant increase in sobriety among patients who either undertook alcoholism therapy, accepted all parts of the study or kept clinical appointments.