Research assessing the relationship of the five-factor model (FFM) to personality disorder symptomatology has generally been confirmatory, with three exceptions. The exceptions have been failures to confirm associations of conscientiousness with the obsessive-compulsive personality disorder, agreeableness with dependent, and openness with schizotypal. Haigler and Widiger demonstrated empirically years ago that this was occurring because the predominant FFM measure at that time, the NEO Personality Inventory-Revised, does not include a sufficient representation of maladaptive variants of the respective FFM personality trait domains. Research since their study has continued to fail to confirm the FFM hypotheses, using other measures of the FFM. The current study extended the work of Haigler and Widiger by considering three additional FFM measures, the Big Five Inventory-2 (BFI-2), the International Item Pool-NEO-120 (IPIP-NEO-120), and the Inventory of Personal Characteristics-5 (IPC-5). Data were obtained from a community sample of adults with experience of mental health treatment. The results confirmed an improvement in the FFM-personality disorder relationships when the experimentally manipulated versions of the BFI-2, IPIP-NEO-120, and IPC-5 were used. The implications of the findings for existing and future FFM-personality disorder research are discussed.