l i r t a P H p n U s p t s e c i t F or more than a century, psychologists and psychiatrists have sought to understand the biological bases of mental disorders, with a view to fashioning more effective treatments. Advances in functional neuroimaging have made it possible for clinical researchers to make significant strides towards realizing this long-held goal, allowing them to do the following: 1) characterize differences in neural responses between patients with various psychiatric conditions and healthy subjects (HS), and 2) examine the neural correlates of psychosocial and biologic treatments. One particular target of investigation has been the anxiety disorders, which constitute the most common family of psychiatric disorders, with a lifetime prevalence of 28.8% (1). Compared with HS, patients with anxiety disorders demonstrate hyperactivity in the insula and amygdala, key regions in the fear network (2,3). In addition to shared features, there are important differences among anxiety disorders. For example, hypoactivation in the rostral anterior cingulate cortex (ACC), dorsal ACC, ventromedial prefrontal cortex (PFC), and thalamus is more characteristic of posttraumatic stress disorder than social anxiety disorder (SAD) or specific phobias (2). In contrast, hyperactivation in the amygdalae and insular cortex is found more frequently in SAD and specific phobias than in posttraumatic stress disorder (2). Related research examining the neural correlates of psychosocial treatments for anxiety disorders has found that cognitive behavioral therapy (CBT) decreases blood flow to the amygdala and the medial temporal lobe in patients with SAD during an anxiety-inducing task (4). Research in patients with specific phobias has demonstrated that CBT leads to reduced activation in the anterior insula and increased activation in the dorsolateral PFC (4). These findings suggest that the neural mechanisms underlying anxiety may be specific to each disorder and that the mechanisms associated with change during CBT may be disorder specific as well. Despite the wealth of research examining the neural correlates of anxiety disorders, surprisingly little is known about the neurobiology of panic disorder (PD). Panic disorder is marked by internally generated threat, vegetative symptoms, and sudden extreme anxiety (3). Maladaptive associative learning is thought to play a role in the etiology of PD, and research has shown that patients with PD extrapolate learned fears to related stimuli (5). Agoraphobia (the avoidance of situations in which escape or assistance may not be possible) commonly co-occurs with panic disorder (PD/A) (6). The neural correlates of PD/A are not well described, but one positron emission tomography study found that, compared with HS, PD/A patients displayed higher levels of glucose uptake in the bilateral amygdala, hippocampus, and thalamus (among other regions) (7), providing evidence for the role of the fear network in PD/A. Only two studies have investigated the neural correlates of CBT for PD/A,