Abstract: Background: Client dropout is commonly used as an indicator of quality and effectiveness of drug treatment. Following increasing cocaine use in recent years, research has attempted to identify predictors of retention in treatment for cocaine users but there is no consensus about how individual characteristics and system variables (referral source, treatment setting), what we term here as the “macro-environment” – effect risk of dropout. This study sought to identify macro-environmental factors and examine the way these impact upon treatment retention.Methods: A retrospective longitudinal study of an admission cohort among drug services in Vasta Romagna in Emilia Romagna Region, Italy (up to 8years of treatment or until discharge) was conducted to determine the effect of macro-environmental variables on retention among first time admitted cocaine-dependent clients in different treatment settings. The sample consisted of 1178 clients meeting DSM-IV-R criteria for cocaine dependence.Results: The joint effect of individual and system factors had a significant impact on dropout rates. In particular, lower rates of dropout were observed for those treated in prison and those who had a stable home, HR: 0.09 (0.02–0.48), or lived in rehabilitation units, HR: 0.36 (0.15–0.88), and among clients referred by the local authority and those living with parents, HR: 0.60 (0.38–0.95).Conclusion: The combined effect of individual and system factors on retention in treatment sets a critical background necessary to assess any impact of organizational dynamics and delineate the trajectory for future interventions.