Abstract Narcissistic personality disorder is the subject of extensive discussion in the literature. Yet, the validity of this diagnostic category remains questionable. This is owed, in large part, to the relative absence of empirical work that has examined narcissism in clinical samples. Descriptions and findings from studies involving non-clinical samples suggest that narcissism is associated with considerable interpersonal impairment. The objective of the present study was to examine this possibility in a sample of psychiatric outpatients. Consecutively admitted patients ( N = 240) to a day treatment program completed measures of narcissism, interpersonal problems, and general psychiatric distress. Patients were categorized into high, moderate, and low narcissism groups. The groups were compared on overall interpersonal impairment, as well as on particular domains of interpersonal behavior. Treatment duration and discharge status were also compared among the three groups. Analysis of covariance and chi-square analyses were used. At baseline, higher levels of narcissism were significantly associated with greater interpersonal impairment. The interpersonal style of the more narcissistic patients was particularly characterized by domineering, vindictive, and intrusive behavior. At post-treatment, only the association between narcissism and intrusive behavior remained significant. Change in interpersonal difficulties following treatment did not differ significantly among the groups. However, failure to complete treatment was associated with narcissism. The results underscore the interpersonal impairment associated with narcissism and support the notion of narcissistic personality disorder as a valid diagnostic category.