Person-reliability indices can assist clinicians in determining the interpretability of a patient's responses to the Basic Personality Inventory (BPI). Using an initial sample of 65 psychiatric patients, we found that: (1) different person-reliability indices showed modest evidence of psychometric adequacy and tended not to be confounded with general psychopathology; (2) a content consistency index of person reliability was predictably related to other item change variables, whereas within-session profile stability was related to across-session measures of profile stability: and (3) evidence for the ability of person-reliability indices to moderate the validity of clinical criteria was weak. Results provide cautious support for a multidimensional conceptualization of the person reliability construct on the BPI but demand further evaluation of the clinical utility of person reliability indices.