Abstract
A group of alcoholic men were asked 8 years postdischarge from hospital if they could drink alcohol with control. Subjects who said “no” (the NCD group) were compared to those who said “yes” (the CD group). A majority of the NCD group had been abstinent during the year preceding the 8-year evaluation, whereas none of the CD group had been. In the model developed, troubles resulting from drinking are seen as instrumental in changes from a CD to a NCD belief. The clinical relevance of our results is discussed.