Pii: S0301-0511(01)00091-6

Although DSM-IV criteria for anxiety disorders include physiological symptoms, these symptoms are evaluated exclusively by verbal report. The current review explores the background for this paradox and tries to demonstrate on theoretical and empirical grounds how it could be resolved, providing new insights about the role of psychophysiological measures in the clinic. The three-systems approach to evaluating anxiety argues that somatic measures as well as verbal and behavioral ones are indispensable. However, the low concordance between these domains of measurement impugns their reliability and validity. We argue that concordance can be improved by examining the relationship of variables less global than anxiety and by restriction to specific anxiety disorders. For example, recent evidence from our and other laboratories indicate a prominent role of self-reported and physiologically measured breathing irregularities in panic disorder. Nonetheless, even within a diagnosis, anxiety patients vary radically in which somatic variables are deviant. Thus, in clinical practice, individual profiles of psychological and physiological anxiety responses may be essential to indicate distinct therapeutic approaches and ways of tracking improvement. Laboratory provocations specific to certain anxiety disorders and advances in ambulatory monitoring vastly expand the scope of self-report and physiological measurement and will likely contribute to a refined assessment of anxiety disorders. © 2001 Elsevier Science B.V. All rights reserved.