Borderline personality disorder (BPD) and posttraumatic stress disorder (PTSD) are commonly co-occurring disorders with comorbidity rates up to 58% (Harned, Rizvi, & Linehan, 2010; Zanarini, Frankenberg, Hennen, Reich, & Silk, 2004; Yen et al., 2002). Individuals with BPD and PTSD are more impaired in a variety of areas than those with either disorder alone (Bolton, Mueser, & Rosenberg, 2006; Harned et al., 2010; Pagura et al., 2010), and PTSD often maintains or exacerbates BPD criterion behaviors such as suicidal and nonsuicidal self-injury (NSSI), other impulsive, self-destructive behaviors (e.g., substance use), emotion dysregulation, and dissociation (for a review see Harned, in press). Accordingly, PTSD has been found to decrease the likelihood of remitting from BPD over 10 years of naturalistic follow-up (Zanarini, Frankenburg, Hennen, Reich, & Silk, 2006). Despite the severity and chronicity of impairment in this population, little research has evaluated effective approaches for treating PTSD among BPD patients, particularly those with a severe level of disorder (e.g., suicidal and self-injuring patients). One approach is to provide treatment focused primarily on PTSD. Although several empirically supported treatments for PTSD exist that have been found effective with less severe BPD patients (Feeny, Zoellner, & Foa, 2002; Clarke, Rizvi, & Resick, 2008), these treatments remain largely inaccessible to patients with severe BPD. Exposure-based treatments such as prolonged exposure (PE; Foa, Hembree, & Rothbaum, 2007) have the most empirical support for PTSD (Institute of Medicine, 2007; Powers, Halpern, Ferenschak, Gillihan, & Foa, 2010), but severe BPD patients are likely to be excluded due to safety concerns (e.g., acute suicidality, recent serious self-injury) or other severe comorbidities (e.g., substance dependence, dissociative disorders; Foa et al., 2007). Clinicians in routine practice often use even broader exclusionary criteria for PE that would rule out most if not all BPD patients, including suicidality, dissociation, depression, a history of multiple childhood traumas, and any comorbid diagnosis (Becker, Zayfert, & Anderson, 2004; van Minnen, Hendriks, & Olff, 2010). Similarly, the consensus among PTSD experts is that a treatment approach based primarily on memory processing (such as PE) is inappropriate for cases of “complex PTSD” (i.e., PTSD with associated features that are common in severe BPD such as emotion dysregulation, behavioral dysregulation, and dissociative symptoms; Cloitre et al., 2011). Given the common practice of excluding severe BPD patients from PTSD treatment, little is known about the efficacy of these treatments for this population. A second approach is to provide treatment focused primarily on BPD. Dialectical behavior therapy (DBT; Linehan, 1993) is the most empirically supported treatment available for BPD and has been found effective in reducing suicide attempts, nonsuicidal self-injury (NSSI), general psychological distress, crisis service use, and treatment dropout (Kliem, Kroger, & Kosfelder, 2010; Leichsenring, Leibing, Kruse, New, & Leweke, 2011). DBT focuses primarily on helping BPD patients achieve behavioral control by increasing behavioral skills, and does not typically directly target PTSD. When PTSD is not directly targeted, the rate of PTSD remission during DBT is relatively low (35%; Harned et al., 2008). Nonetheless, treatments such as DBT that focus on teaching coping skills (e.g., emotion regulation, interpersonal effectiveness) are viewed by PTSD experts as safer and more appropriate than treatments focused on memory processing for patients with “complex PTSD” (Cloitre et al., 2011). Thus, DBT may be a reasonable and effective alternative to PTSD-focused treatment for severe BPD patients with PTSD. A third approach is to provide an integrated treatment that targets BPD and PTSD simultaneously. In this case, integrating DBT and PE would represent the best evidence synthesis of the available treatments for BPD and PTSD, respectively. In addition, DBT has been shown to be effective in reducing behaviors commonly used as exclusion criteria for PE among severe BPD patients with PTSD, including imminent suicide risk, suicide attempts, NSSI, severe dissociation, and substance dependence (Harned, Jackson, Comtois, & Linehan, 2010). Thus, an integrated treatment approach that utilizes DBT to increase behavioral skills and achieve stabilization prior to and during PE may be optimal for severe BPD patients with PTSD, and a recent open trial of an integrated DBT and PE treatment showed promising results (Harned, Korslund, Foa, & Linehan, 2012). This type of phase-based treatment (coping skills followed by memory processing) is also endorsed by a majority of PTSD experts as a frontline treatment for “complex PTSD” (Cloitre et al., 2011). Although clinicians and PTSD experts generally believe that PE alone is inappropriate for severe BPD patients and that a skills-focused treatment such as DBT, possibly combined with PE, is preferable, it is unknown whether severe BPD patients share these beliefs. Current best practice standards emphasize the importance of considering patient preferences in the treatment decision-making process (American Psychological Association, 2006), and research indicates that providing patients with their preferred treatment improves outcomes and reduces dropout (Swift & Callahan, 2009). Studies of treatment preferences for PTSD have found that exposure therapy, including PE, is a preferred treatment in undergraduate, trauma-exposed, and treatment-seeking PTSD samples (Angelo, Miller, Zoellner, & Feeny, 2008; Becker, Darius, & Schaumberg, 2007; Cochran, Pruitt, Fukuda, Zoellner, & Feeney, 2008; Feeny, Zoellner, Mavissakalian, & Roy-Byrne, 2009; Tarrier, Liversidge, & Gregg, 2006; Zoellner, Feeny, & Bittinger, 2009; Zoellner, Feeny, Cochran, & Pruitt, 2003), and preliminary evidence suggests that providing PTSD patients with their preferred treatment enhances outcomes (Feeny et al., 2009). These findings suggest that the underutilization of exposure treatments for PTSD, including PE, appears to be primarily due to therapist factors (e.g., lack of training, concerns about PE) rather than client factors (e.g., unwillingness to participate in PE). Several of these studies have also begun to examine predictors of treatment preference, finding that demographic factors (higher education, nonminority status) predict a preference for PE over sertraline, whereas psychopathology factors (more severe PTSD, depression, and anxiety) predict a preference for sertraline over PE (Angelo et al., 2009; Feeny et al., 2009; Zoellner et al., 2009). To date, no research has examined treatment preferences among BPD patients with PTSD specifically, and understanding the treatment preferences of these patients may help to inform and enhance treatment for this difficult-to-treat population. The present study examines this issue in a treatment-seeking sample of suicidal and self-injuring women with BPD and PTSD. Our three primary aims were to (a) determine whether suicidal and self-injuring BPD women with PTSD prefer DBT, PE, or a combined DBT and PE treatment, (b) evaluate the reasons underlying treatment preference, and (c) examine potential predictors of treatment preference that have been evaluated in prior studies (i.e., demographics, PTSD severity, psychological distress/comorbidity) as well as ones new to this study and client population (i.e., intentional self-injury history, emotional experiencing).