The influence of cannabis abuse on the severity of existing psychotic and affective symptoms is still unclear. Among 470 consecutively admitted psychotic or affective patients, 54 active (in the previous month) cannabis abusers were detected via urine tests (Sure Step TM kits; Applied Biotech Inc, San Diego, Calif) and Structured Clinical Interview for DSM-IV (SCID- IV) questionnaire. In 24 cases, substances other than cannabis were abused; 392 patients were nonabusers. All patients were diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria. The following rating scales were used: Hamilton Depression Rating Scale (HAM-D-21), Positive and Negative Syndrome Scale (PANSS), and Young Mania Rating Scale (YMRS). Cannabis abusers (n = 54) were significantly younger and more frequently males than nonuser patients. In this group, there were more schizophrenic patients and fewer affective and anxiety patients (chi(2) = 11.76; P < .01). The double-diagnosed patients had more prominent psychotic symptoms than the nonusers (n = 392)-PANSS positive: 19.056 +/- 8.30 vs 16.128 +/- 8.031 (P < .02; t(446) = 2.510). The difference was statistically significant for hallucinatory behavior, excitement, grandiosity, and hostility. General PANSS scale rate of abusers was lower: 33.012 +/- 9.317 vs 37.3575 +/- 11.196 (P < .01; t = 2.727), especially for depression, anxiety, somatic concern, guilt feelings, tension, motor retardation, and volition disturbances. Rates of PANSS negative scale of abusers and nonusers were not significantly different (13.815 +/- 6.868 vs 14.983 +/- 6.446) except for lower rates of social withdrawal and stereotyped thinking for abusers. No significant difference in general level of manic symptoms (YMRS) between abusers and nonusers was observed (6.778 +/- 10.826 vs 4.910 +/- 7.754), but severity of thought/language disturbances and poor insight was found significantly higher in the abusers. Cannabis abusers are obviously less depressive (HAM-D): 5.944 +/- 10.291 vs 12.896 +/- 13.946 (P < .0005, t = 3.535). Such differences were observed in the high number of the subscales. Abusers' rates were higher (although not significantly) for paranoid symptoms and general somatic symptoms. Cannabis possibly produces some antidepressive and anxiolytic effect on psychotic and affective inpatients. The "price" of this effect is often an exacerbation of psychotic and some manic symptoms.