INTRODUCTION: Although cannabis use may be involved in the aetiology of acute psychosis, there has been considerable debate about the association observed between cannabis use and chronic psychosis. In particular, because of the frequent co-occurrence between schizophrenia and cannabis use, the question has been raised of a causal link between exposure to cannabis as a risk factor and the development of psychosis or psychotic symptoms. OBJECTIVE: The aim of this article was to examine the evidence that cannabis use causes chronic psychotic disorders by using established criteria of causality. These criteria were defined by: biologic plausibility, strength of the interaction between the risk factor and the disease, reprieability of the results, temporal sequence between the exposure to the risk factor and the beginning of the disease and existence of a dose-effect relationship. METHODS: The selected studies were found in Medline using the keywords "cannabis" and "psychosis", "cannabis" and "schizophrenia", "cannabis" and "psychotic symptoms" and "prospective" or "cohort" or "longitudinal". The selected studies were all prospective studies assessing the temporal sequence between cannabis use and emergence of psychosis or psychotic symptoms. The search strategies resulted in 60 records that were screened by reading both titles and abstracts. Seventeen studies were considered eligible, and then, after reading the full text, seven met the inclusion criteria. RESULTS: Together, the seven studies were all prospective cohorts and represented 50,275 human subjects. There were three European studies (from Sweden, Holland and Germany), one from New Zealand and one from Australia. Only one study of the seven did not show a significant association between cannabis consumption and increase of the risk of developing a psychosis. However, this study had some bias, such as low level of cannabis use and the lack of evaluation of cannabis use after inclusion. For the six other studies, data show the existence of a significant association between cannabis use and psychotic disorders (with an increased risk between 1.2 and 2.8 in Zammit et al.'s study), particularly among vulnerable individuals (that is with a prepsychotic state at the time of inclusion). Therefore, all the studies that assessed a dose-effect relationship showed this link between cannabis use and the emergence of psychosis or psychotic symptoms. The fact that all causal criteria were present in the studies suggests that cannabis use may be an independent risk factor for the development of psychosis. Results seem to be more consistent for vulnerable individuals with the hypothesis that cannabis use may precipitate psychosis, notably among vulnerable subjects. In particular, early onset of cannabis use during adolescence should be an environmental stressor that interacts with a genetic predisposition to induce a psychotic disorder. CONCLUSION: The objective of this article was to examine whether cannabis use can be an independent risk factor for chronic psychotic disorders, by using established criteria of causality. Data extracted from the selected studies showed that cannabis use may be an independent risk factor for the development of psychotic disorders. Early screening of the vulnerability to psychotic disorder should permit improved focus on prevention and information about the specific risks related to cannabis use among this population.