Background: some addictovigilance studies describe off-label use of morphine sulfate in non-chronic pain opioid-dependent patients for recreational purposes or as an alternative to conventional substitution treatments. The objective of this study is to provide comprehensive data on the risks associated with this off-label use. Methods: a retrospective cohort of non-chronic pain patients who benefited from regular dispensation of morphine sulfate between April 1, 2012 and December 31, 2014 was based on the comprehensive health insurance database (SNIIRAM). Sociodemographic characteristics and prevalence of associated comorbidities were investigated. The incidence of overdoses, deaths, doctor shopping behavior and complications (i.e. viral and bacterial infections and thrombotic complications) related to the very frequent practice of intravenous administration were compared by the Cox regression method to those treated by opioid substitution therapy (OST): buprenorphine or methadone). Results: 1075 non-chronic pain opioid-dependent patients were included (76.6% of men, median age: 33 years). Overdose incidences were significantly higher than OST patients. Except for thrombotic complications versus OST and hepatitis C virus versus methadone group, incidence of the other associated risks was significantly higher in morphine sulfate group. Conclusion: This study is the first to assess the risks associated with off-label use of morphine sulfate at national level. It questions the relevance of these prescriptions as a secure opioid maintenance therapy.