This study compared the manifest dream content of 20 schizophrenic adolescent inpatients whose medications were stable for at least four weeks, 21 adolescent inpatients with other mental disorders (nonschizophrenic group) matched for age and gender, and 31 matched community controls. All participants were administered the standardized Formal Dream Content Rating Scale (FDCRS), which evaluates dream-related anxiety, cognitive disturbance, implausibility, involvement, primitivity, and recall, as well as two additional scales measuring emotional expression and duration of dream report. The Positive and Negative Symptoms Scale (PANSS) was administered to the two inpatient groups. The community controls demonstrated more involvement and emotional expression than the schizophrenic patients; furthermore, they demonstrated more implausibility and had a greater duration of dream report compared with the nonschizophrenic group. In the schizophrenic patients only, elevated scores on the negative subscale of the PANSS were significantly correlated with lower scores on involvement, emotional expression, and dream recall. No relationship was found between the positive subscale of the PANSS and any of the FDCRS subscales. These results suggest that psychopathology per se, rather than the specific psychiatric disturbance, may be associated with impoverishment of dream content, and that negative, rather than positive, schizophrenic symptomatology may be influential in the dream content of schizophrenic youngsters.