Abstract In a sample of 412 patients with psychotic disorders, the authors examined whether comorbid substance use can be reliably diagnosed, is associated with increased rates of affective symptoms and syndromes and specific psychotic symptoms, and is associated with lowered reliability of the DSM-III-R principal diagnosis. Data from the DSM-IV Field Trial for Schizophrenia and Other Psychotic Disorders was analyzed. In this dataset, substance use was scored on a 4-point ordinal scale and reliability was determined using weighted kappa scores. The associations of significant substance use with affective syndromes and symptoms, and psychotic symptoms were analyzed. Kappa statistics were calculated for principal psychotic disorder diagnoses for patients with and without significant substance use. Weighted kappa scores for substance use ratings ranged from 0.27 to 0.96 (median = 0.85). Syndromal depression was significantly associated with current alcohol use in the entire sample and in the subgroup with schizophrenia alone. Grandiose delusions were also associated with substance use. Significant comorbid substance use was not associated with lowered reliability of diagnosing the principal psychotic disorder. These findings support the hypothesis that comorbid substance abuse can be reliably diagnosed and that alcohol abuse is associated with depressive syndromes in patients with psychotic disorders.