The Machiavellian Personality Scale (MPS; Dahling, Whitaker, & Levy, 2009) was designed to measure four content subdomains (i.e., amorality, desire for control, desire for status, and distrust of others). In this study, we evaluated the structural validity of MPS by testing a series of six competing models, including the one-factor CFA model, the four-factor CFA model, the higher-order CFA model, the exploratory structural equation modeling (ESEM) model, the bifactor CFA model, and the bifactor exploratory structural equation modeling (B-ESEM) model. The findings suggested that the B-ESEM model, which took two sources of psychometric multi-dimensionality into account, provided the best fit to the data. Although the Machiavellian Personality Scale was dominated by the global Machiavellianism, the measurement quality of the subdomains varied, with the desire for control sub-factor being more clearly defined than the other three specific factors (i.e., amorality, desire for status, and distrust of others). Recommendations were made about the use of the Machiavellianism Personality Scale in practice. Ethical statement: All the subjects gave written informed consent, and the protocol was approved by the research ethics review board of Xinyang Normal University. The authors have declared that no competing interests exist. (C) 2016 Elsevier Ltd. All rights reserved.