Despite scant evidence of an advantage of combining medication and cognitive behavioral therapy (CBT) for social anxiety disorder, the advantages and disadvantages of the two approaches suggest the possibility that some form of combination may be helpful for many clients. At the same time, the available evidence and theoretical considerations suggest that some methods of combination could provide short-term benefits but long-term decreases in efficacy compared to either treatment alone. Thus, current knowledge suggests that, for social anxiety disorder, more treatment is no substitute for well-informed treatment. We provide guidelines for treatment based on available evidence, theoretical models, and clinical experience. Overall, the clearest case for combination treatment is for the addition of an empirically supported CBT for social anxiety disorder following the stabilization of response to an empirically supported psychopharmacological agent. The CBT should then be modified to address the use of medication, the possibility of relapse, and, ideally, the tapering of medication during CBT. Further research is called for in order to support these recommendations and extend the available literature. Keywords: social anxiety disorder; social phobia; treatment; psychotherapy; medication; CBT A review of the extant treatments for social anxiety disorder makes it dear that multiple psychological and pharmacological interventions produce beneficial effects (e.g., Rodebaugh, Holaway, & Heimberg, 2004) and that the psychotherapies that produce these effects are primarily forms of cognitive behavioral therapy (CBT). Both research and clinical practice are therefore not only concerned with the question of what works at all, which occupied the field for much of the past, but also with the additional question of how we can maximize the impact of generally effective techniques. In considering this question, it is vital to note that both medications and CBT have advantages and disadvantages. At least some research (e.g., Davidson et al., 2004; Heimberg et al., 1998) and general clinical opinion indicate that medications may work more quickly. However, medications also appear to be associated with significant rates of relapse (e.g., 30% to 60% for paroxetine; e.g., Stein, Veriani, Hair, & Kumar, 2002; Stein et al., 1996) that are in excess of those reported for CBT (e.g., 17% for group CBT; liebowitz et al., 1999). Conversely, CBT is typically thought to take more time but is associated with lower relapse rates. Thus, although there is no guarantee that any combination of a given medication and a given form of CBT will increase response or decrease relapse, it appears logical to explore their combination further. In addition, and more pragmatically, many clients requesting psychotherapy are already on a medication. For example, we recently completed a study of CBT in collaboration with Debra Hope of the University of Nebraska-Lincoln in which participants were permitted to stay on a stable dose of their current medication (Zaider, Heimberg, Roth, Hope, & Turk, 2003). Over half of the participants were on some form of psychoactive medication when they presented to one of the participating clinics. We have no reason to believe that these clients were substantively different from those presenting at other psychotherapy clinics, at least in terms of medication use. It is likely that the percentage of persons with social anxiety disorder who present for psychotherapy already on medication has increased since the beginning of direct-to-consumer advertising of medications for social anxiety disorder a few years ago. Therefore, psychotherapists may be increasingly forced to deal with the issue of combining treatments if they wish to treat the typical client with social anxiety disorder. This article reviews the current state of knowledge about combination treatments for social anxiety disorder and provides guidelines for the practicing clinician regarding combination treatments. …