Transcranial magnetic stimulation (TMS) has been applied to a growing number of psychiatric disorders as a neurophysiological probe, a primary brain-mapping tool, and a candidate treatment. Although most investigations have focused on the treatment of major depression, increasing attention has been paid to anxiety disorders. The aim of this study is to summarize published findings about the application of TMS as a putative treatment for anxiety disorders. TMS neurophysiological and mapping findings, both clinical and preclinical, have been included when relevant. We searched Medline, PsycInfo, and the Cochrane Library from 1980 to January 2009 for the terms ‘generalized anxiety disorder’, ‘social anxiety disorder’, ‘social phobia’, ‘panic’, ‘anxiety’, or ‘posttraumatic stress disorder’ in combination with ‘TMS’, ‘cortex excitability’, ‘rTMS’, ‘motor threshold’, ‘motor evoked potential’, ‘cortical silent period’, ‘intracortical inhibition’, ‘neuroimaging’, or ‘intracortical facilitation’. Most of the therapeutic experiences with repetitive TMS available in the literature are in the form of case reports, not controlled or blinded studies. Stimulation of the right dorsolateral prefrontal cortex, especially at high frequencies, has been reported to reduce anxiety symptoms in posttraumatic stress disorder and panic disorder; nevertheless, results are mixed. A specific role for the right dorsolateral prefrontal cortex in the posttraumatic stress disorder symptom core can be hypothesized. TMS remains an investigational intervention that has not yet gained approval for the clinical treatment of any anxiety disorder. Clinical sham-controlled trials are scarce. Many of these trials have supported the idea that TMS has a significant effect, but in some studies, the effect is small and short lived. The neurobiological correlates suggest possible efficacy for the treatment of social anxiety that still has to be investigated.