The two objectives of this tri-ethnic study were (1) to test competing hypotheses from the minority status and ethnic culture perspectives in examining cross-ethnic symptom differences in schizophrenia and (2) to test cultural mediators of the symptom differences. Analyses were done on samples of minority (African-American and Latino) and nonminority (white) groups. Hypothesized cross-ethnic symptom differences were tested, and indicators of sociocentricity were examined as cultural mediators of symptom differences. The sample consisted of 184 individuals (51.6% white, 32.6% African-American, 15.8% Latino, 75% male) diagnosed with schizophrenia in an outpatient, urban setting. Symptom variables were obtained from the Brief Psychiatric Rating Scale and the Quality of Life Scale. Two sociocentric indicators (empathy and social competence) were used to differentiate minority from nonminority groups. Multiple regression was used to test the mediational influence of the sociocentric indicators on the symptom differences. After controlling for social class, significant differences were found in eight symptom variables. These showed the nonminority group to be consistently more symptomatic than the ethnic minority groups, findings which supported the ethnic culture hypothesis. Sociocentric indicators were found to be significant mediators of the cross-ethnic symptom differences. This study supported the ethnic culture hypothesis, which predicted lower symptoms for the ethnic minority groups, and showed that sociocentric variables strongly mediated the more benign symptom profile for the ethnic minority groups. The study indicates that culture should be more fully integrated into current biopsychosocial models of schizophrenia.