Background: Normal and pathological narcissism have been the focus of considerable theoretical discussion and empirical research in recent years in personality psychology and psychopathology. Kernberg [1-4] has argued that there is a particularly dysfunctional and impairing variant of narcissistic disturbance known as malignant narcissism. This exploratory study sought to develop, using established assessment methods, a dimensional measure of malignant narcissism that incorporates the key features of grandiose narcissism, paranoid propensities, psychopathic features, and proclivity for a sadistic and aggressive interpersonal style. Method and Sampling: This study examined 57 subjects, diagnosed with borderline personality disorder that were treated using 3 different empirically supported treatments in a previous study , for possible deviance on the proposed malignant narcissism index. It also evaluated 2 important clinical domains of change in relation to malignant narcissism. To wit, it was predicted, based on Kernberg’s [3, 4] clinical model, that elevated levels of malignant narcissism would be significantly associated with slower rates of improvement in both general psychosocial/psychological functioning and anxiety among treated individuals. Results: Higher levels of malignant narcissism were associated, as predicted, with slower rates of improvement in both global functioning and anxiety. The proposed malignant narcissism index was a more powerful predictor of slowed improvement in global functioning than simple narcissistic personality disorder features. Conclusions: The heuristic potential of the malignant narcissism construct is discussed and the utility of a dimensional approach to this construct is explored, especially in reference employing personality traits/processes to better understand pathological configurations and personality disturbance.