The Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT) is a key measure of social cognition recommended by the MATRICS committee. While the psychometric properties of the MSCEIT appear strong, previous evidence suggested its factor structure may have shifted when applied to schizophrenia patients, posing important implications for cross-group comparisons. Using multi-group confirmatory factor analysis, we explicitly tested the factorial invariance of the MSCEIT across schizophrenia (n=64) and two normative samples (n=2099 and 451). Results indicated that the factor structure of the MSCEIT was significantly different between the schizophrenia and normative samples. Implications for future research are discussed.