Social anxiety disorder (SAD) is among the most common anxiety disorders (Kessler et al., 2005). Epidemiological surveys further show that SAD, and particularly the generalized subtype, is associated with significant personal and societal costs (Schneier et al., 1994; Schneier, Johnson, Hornig, Liebowitz, & Weissman, 1992), underscoring the need for effective interventions. Cognitive-behavioral therapy (CBT) has been established as an efficacious intervention for SAD (Hofmann & Smits, 2008). CBT emphasizes repeated or prolonged confrontation to feared cues often in combination with cognitive restructuring (Hofmann & Otto, 2008), with the goal of assisting patients in reacquiring a sense of safety in social interaction and performance situations (Otto, Smits, & Reese, 2004). CBT, delivered in individual or group format, has been shown to outperform both pill and psychological placebo treatment (Davidson et al., 2004; Heimberg et al., 1998) and yields medium to large reductions in symptom severity (Clark et al., 2003). Although efficacious, CBT results in suboptimal outcomes for many. Indeed, treatment non-response rates in large clinical trials have been 50% or higher (Davidson et al., 2004), and most individuals do not achieve remission (Blanco et al., 2010; Davidson et al., 2004). The observation of suboptimal outcomes of CBT for SAD has prompted research on augmentation strategies. One exciting development in this area has been the application of dcycloserine (DCS), a partial agonist at the N-methyl-D-aspartate (NMDA) receptor (Anderson & Insel, 2006; Davis, Ressler, Rothbaum, & Richardson, 2006; Hofmann, Smits, Asnaani, Gutner, & Otto, 2011). Following preclinical research demonstrating that DCS can facilitate extinction learning in rodents by enhancing NMDA receptor function (Davis et al., 2006), initial studies of DCS for augmenting exposure therapy for the anxiety disorders have yielded promising results (Bontempo, Panza, & Bloch, 2012). With respect to SAD, Hofmann and colleagues (Hofmann, Meuret, et al., 2006) initially showed that patients receiving DCS 1 hour prior to the last 4 sessions of a 5-session CBT protocol evidenced significantly greater rates of improvement than patients who received pill placebo. This finding was later replicated by an independent group (Guastella, 2008). Given these positive findings, DCS as an augmentation strategy was subsequently applied to a standard 12-session course of group CBT for SAD, examining the degree to which it extends treatment gains in a full rather than an abbreviated protocol (Hofmann et al., in press) As has been the case for CBT as a single modality treatment, response to DCS-augmented CBT has been variable across individuals with SAD (Hofmann, Pollack, & Otto, 2006; Hofmann et al., in press). The aim of the present study is to identify individual characteristics that differentiate response to DCS augmentation in a large-scale clinical trial comparing DCS-augmented group CBT to placebo-augmented group CBT in medication-free adults with generalized SAD (Hofmann et al., in press). In the context of this study, individual characteristics associated with differences in outcome between the experimental and the control condition are considered moderators (Kraemer, Wilson, Fairburn, & Agras, 2002) or prescriptive variables (Fournier et al., 2009), whereas individual characteristics associated with outcome regardless of study condition are considered predictor or prognostic variables (Fournier et al., 2009). Accordingly, in addition to identifying individual characteristics that differentiate responding to DCS, this analysis has the potential to identify individual characteristics that are prognostic of responding to CBT for SAD regardless of DCS augmentation. We hope that the findings of this study will stimulate research aiming to refine the understanding of the mechanisms by which CBT (with or without DCS) reduces anxiety pathology, and thereby aid in achieving the goal of personalized therapy (Anderson & Insel, 2006). Constrained by the limitations of the assessment protocol for the parent trial, we evaluated the patient characteristics available in our database that are also routinely assessed in clinical practice (e.g., demographic characteristics, clinical characteristics, and personality traits). Research relating the efficacy of DCS augmentation to these patient characteristics is limited to some preliminary work in PTSD. De Kleine and colleagues (de Kleine, Hendriks, Kusters, Broekman, & van Minnen, 2012) recently reported that differences between DCS-augmented CBT and placebo-augmented CBT with respect to symptom reduction were greater among patients who presented with more severe symptoms than patients who presented with less severe symptoms at baseline. However, because this post-hoc moderator analysis did not include any covariates, it is impossible to rule out possible third-variable explanations for the clinical severity-DCS efficacy relation observed in this study (e.g., comorbid diagnoses, sex, neuroticism, etc.). Extant research on the relation between patient characteristics and CBT outcome for SAD has frequently pointed to depression severity as an important predictor (i.e., prognostic factor), although the findings have been mixed. Specifically, baseline depression severity has been associated with greater symptom severity at the end of CBT for SAD (Collimore & Rector, 2012; Erwin, Heimberg, Juster, & Mindlin, 2002; Ledley et al., 2005; Turner, Beidel, Wolff, Spaulding, & Jacob, 1996) and with less symptom improvement during CBT (Chambless, Tran, & Glass, 1997; Ledley et al., 2005). Several studies have, however, failed to observe the relation between depression and rate of change during CBT for SAD (Erwin et al., 2002; Joormann, Kosfelder, & Schulte, 2005; Turner et al., 1996), suggesting that baseline depression severity may simply be associated with greater symptom severity before and after treatment, but not necessarily the efficacy of CBT for SAD. Interestingly, McEvoy and colleagues (McEvoy, Nathan, Rapee, & Campbell, 2012) did not identify comorbid depression as a predictor of CBT for SAD outcome in an analysis that also included a number of other individual characteristic variables (e.g., age, gender, social anxiety symptom severity, number of comorbid diagnosis, degree of life interference). Instead, social anxiety symptom severity and number of comorbid diagnosis emerged as prognostic factors in this analysis. Together, these findings highlight the importance of evaluating multiple prospective prognostic or prescriptive variables simultaneously rather than in isolation. Although it has been proposed that personality traits have implications for the outcome of psychotherapy for anxiety disorders (Zinbarg, Uliaszek, & Adler, 2008), there has been no research to our knowledge evaluating the prognostic (or prescriptive) value of personality traits (e.g., agreeableness, conscientiousness, extraversion, neuroticism, and openness) with respect to the outcome of CBT for SAD (or DCS augmentation). Given findings in the literature linking conscientiousness to health behaviors and greater treatment adherence (Axelsson, Brink, Lundgren, & Lotvall, 2011; Hill & Roberts, 2011), it is plausible that those low in conscientiousness are less able to comply with or utilize treatment. Because it is notable that the advantages for DCS augmentation are much greater in very brief treatments (Guastella, 2008; Hofmann, Meuret, et al., 2006) than in the current 12-session protocol (Hofmann et al., in press), it is possible that patients low in conscientiousness, who may not receive the full dose of CBT, may benefit more from enhancement with DCS. Should it be the case that the DCS augmentation effects are indeed stronger for individuals for whom a standard course of CBT leaves ample room for improvement (see de Kleine et al., 2012), high agreeableness may also emerge as a prognostic and prescriptive variable in our analysis, as it has been associated with suboptimal responding to exposure-based intervention for panic disorder (Harcourt, Kirkby, Daniels, & Montgomery, 1998). Because extant research on predictors of the efficacy of CBT for SAD and DCS augmentation of CBT, respectively, is limited in scope and has yielded inconsistent findings, we employed an exploratory, hypothesis-generating, rather than hypothesis-driven framework in the present study. Here, we specifically aimed to determine the unique prescriptive and/or prognostic value of each patient characteristic, by evaluating its predictive value while controlling for the influence of other patient characteristics. Acknowledging that multiple tests increase the risk of Type 1 error, we adopted an approach to handling this concern that was developed by Fournier and colleagues (Fournier et al., 2009) and later used by Amir and colleagues (Amir, Taylor, & Donohue, 2011). Specifically, we selected the patient characteristics assessed at the baseline visit and categorized these in meaningful domains (e.g., demographic characteristics, clinical characteristics, and personality traits). The first step involved a step-wise procedure to identify significant prognostic and prescriptive variables within each domain. The second step involved the simultaneous entry of these prescriptive and prognostic variables identified for each domain in one final model.