Using self- and observer reports on the Personality Inventory for DSM-5 (PID-5) and the HEXACO Personality Inventory-Revised (HEXACO-PI-R), we identified for each inventory several trait dimensions (each defined by both self- and observer reports on the facet-level scales belonging to the same domain) and 2 source dimensions (each defined by self-reports or by observer reports, respectively, on all facet-level scales). Results (N = 217) showed that the source dimensions of the PID-5 were very large (much larger than those of the HEXACO-PI-R), and suggest that self-report (or observer report) response styles substantially inflate the intercorrelations and the alpha reliabilities of the PID-5 scales. We discuss the meaning and the implications of the large PID-5 source components, and we suggest some methods of controlling their influence.