A prospective, multicentre study was designed to evaluate the impact of a treatment programme in alcohol-dependent patients (n = 850). Drinking status and measures related to quality of life such as morbidity, mortality, alcohol-related problems, and psychological and social functioning were assessed 5 years from the initial contact with the patient. Five years after the initial visit, drinking status followed a bimodal distribution: 371 patients (43.6%) were found to be abstinent, 323 (38%) were heavy drinkers, and only 55 (6.5%) were controlled drinkers. A total of 65 (7.6%) patients died during the study, 33 (3.9%) patients were lost to follow-up and three (0.3%) patients were not collaborative. Differences in health-related issues were identified for each of the drinking status categories. Abstinent patients presented with significantly better outcomes than controlled and heavy drinkers on most medical, socio-economic and psychological measures. Heavy drinkers showed significantly worse outcomes than controlled drinkers.