Cannabis consuming schizophrenic patients are younger at onset, are likely to have started abuse before onset of schizophrenia and show more prominent positive symptoms than nonabusers. It has been suggested that cannabis is a risk-factor for schizophrenia. Our aim was to assess prevalence and pattern of cannabis use in 125 chronic male schizophrenic subjects and its impact on socioepidemiological and clinical variables as well as which disorder precedes the other in onset. Assessment of consumption was made with a semi-structured clinical interview. Clinical status was assessed by means of the SANS, SAPS, PANSS and BPRS scales. Cannabis consumption was found in 54 subjects (43%), 66.7% of whom started it at least three years before onset of schizophrenia. Consumers were younger and with lower negative symptoms, specially abusers and polysubstance abusers. Family history positive for psychosis was more frequent in consumers, especially when consumption started before onset of schizophrenia. Subjects whose onset of schizophrenia preceded the beginning of cannabis abuse had more positive symptoms than those who started abuse before the onset of schizophrenia. On these grounds, our sample could be subdivided into two main groups, one that uses substances to counter distressing symptoms of schizophrenia and another in which cannabis might be one of the factors predisposing to the disease; the former had less negative symptoms than nonabusers. Our data support both heterogeneity of schizophrenia and genetic susceptibility to environmental agents.