Purpose Political skill is measured with the political skill inventory (PSI), and the construct is composed of four distinct dimensions. Previous validation studies of the PSI found evidence in support of the four-factor structure, but only using self-reports. Furthermore, no efforts have been made to also identify a single, higher-order factor solution through second-order factor analysis. The present research aims to expand on prior work and report on a two-study investigation of both the construct validity and antecedents and consequences of the political skill construct. Design/methodology/approach To test construct validity, Study 1 combined self- and other reports of political skill from 467 employees in a confirmatory factor analysis. Study 2 used longitudinal data from 202 employees to constructively replicate Study 1 results and to test hypotheses regarding the antecedents and consequences of political skill. Findings The results of Study 1 confirmed both a four-factor and a single higher-order factor solution of the political skill construct, thus supporting our hypothesis. Study 2 constructively replicated the Study 1 factorial validity results, and supported hypotheses regarding the dispositional and developmental experience antecedents, career-related consequences, and mediation of these antecedents and outcomes by political skill. Originality/value These two studies test the construct validity of political skill using both self- and other-reports. Further, this is the first research to test the Ferris et al. conceptualization of political skill, by examining its antecedents, consequences, and mediation of the antecedents-consequences relationships.