In treated HIV-infected patients, mortality is now dominated by non-AIDS-related causes in which tobacco smoking is a predominant risk factor. The implementation of tobacco smoking cessation programs is therefore warranted to increase survival but should consider the specificities of this population to be successful. All outpatients consulting in May to June 2004 within the ANRS CO3 Aquitaine Cohort of HIV-infected patients were asked to complete a self-administered questionnaire including questions about tobacco and other drugs consumption, the Fagerstr?est for Nicotine Dependence (FTND), a visual scale to estimate motivation to stop smoking and the Center for Epidemiologic Studies Depression (CESD) scale. Among 509 patients included, mean age was 44 years, 74% were men, 19% were infected through injection drug use, and 257 (51%) were regular smokers (at least one cigarette per day). Among them, 60% had a medium or strong nicotine dependence (FTND = 5), 40% were motivated to quit smoking and 70% had already tried at least once. An FTND of 5 or more was more frequently reported in the 146 smokers (62%) with depressive symptoms compared to other smokers (70% versus 48%). Fifty-five regular smokers (23%) were codependent on cannabis and 31 (12%) to alcohol. Overall, only 35 (14%) regular smokers were motivated, non-codependent, without depressive symptoms, and could be proposed a standard tobacco cessation program. Depressive symptoms were highly prevalent in this representative population of HIV-infected patients. To be successful, smoking cessation interventions should be specifically built to take into account depression and codependencies in addition to nicotine dependence and motivation.