An interviewer rating scale for the measurement of (agora)phobic avoidance, the Phobic Avoidance Rating Scale (PARS), is described and psychometrically evaluated. The criteria of avoidance used are purported to correct weaknesses of the sets of criteria used in self-report instruments. Two samples with inpatients satisfying DSM-III-R criteria for unipolar depressive and/or anxiety disorder were studied. Intercorrelational and factor analyses revealed 3 dimensions among agoraphobic fears: separation fears, social fears and simple phobic fears. Overall, the 3 subscales constructed from these dimensions proved to have good reliability, validity, and sensitivity to change following treatment. The internal consistency of the social subscale seemed to be somewhat low. The PARS should be well suited to assess avoidance among agoraphobic patients on separate dimensions and to distinguish patients with agoraphobia from other psychiatric patients. It seems to be the first interviewer rated agoraphobic scale that is psychometrically investigated and should ideally be used for research purposes in combination with a self-report instrument.