By means of the Stockholm County inpatient care register we identified all cases treated with a diagnosis of cannabis dependence and psychosis, not necessarily at the same occasion, during 1971-1983. By scrutinizing medical records, we evaluated the diagnosis according to DSM-III-R and we assessed the history of substance abuse as well as the psychiatric history and clinical course. We identified 229 cases during the follow-up; 112 of these cases (49%) fulfilled the DSM-III-R criteria for schizophrenia. The majority of the schizophrenics had prominent positive symptoms and a sudden onset of disease, and 69% of the cases had a record of heavy cannabis abuse at least 1 year before onset of psychotic symptoms. The high number of verified DSM-III-R cases of schizophrenia in this cohort and the temporal relation between cannabis abuse and schizophrenia further support the hypothesis that cannabis abuse may be a risk factor for schizophrenia. We confirmed previous observations that cannabis-associated schizophrenia often has a sudden onset and prominent positive symptoms.