Co-morbid substance use and abuse is common in schizophrenic patients, and the role of substance abuse in initiating and maintaining psychosis has important definitional and aetiological implications. We investigated the issue in a cohort of 131 schizophrenic patients. We found non-users (N = 67) were similar to pathological users (N = 64) in current symptomatology and clinical history. The pathological users did, however, have better pre-morbid adjustment levels. Only alcohol use and to some extent cannabis use contributed to this effect; use of stimulants or hallucinogens did not. These results indicate the importance of evaluating the various types of substance used when attempting to explore the significance of co-morbidity. The results also suggest that co-morbidity of substance abuse and schizophrenia may be explained by a common factor antecedent to both: better pre-morbid adjustment. A two-stage model is proposed to explain these findings: increased sociability increases exposure to opportunities of substance use in a subset of patients; subsequent onset of psychotic illness accelerates the use to a pathological level as the individual attempts to cope with the stress of the developing mental illness.