Compulsory treatment for alcoholism and drug addiction exists in many countries, but has been less evaluated. For legal and ethical reasons, controlled designs are scarce. The Swedish �Law on Care for Addicts in certain cases� was implemented in 1982, and revised 1989. Since 1982, twenty outcome studies using different research design have been published, mostly in Swedish. The presentation reviews these studies. Outcomes of men vs. women, alcoholics vs. drug addicts and compulsorily vs. voluntarily treated are compared. Three studies analysed data using multivariate regression models in order to predict programme retention, absconding, substance use outcome, participation in self help groups post discharge and long term mortality. Long term mortality is also analysed using survival models (Kaplan-Meier) and standardised mortality ratios.