Panic disorder with or without agoraphobia is a prevalent clinical disorder which places heavy demands on treatment resources in primary care. The efficiency of delivery of psychological treatments for this disorder is therefore important. Previous research has focused on psychological treatments delivered with reduced therapist contact but methodological problems preclude firm conclusions. The present study investigated the relevance of therapist contact in cognitive behaviour therapy for panic disorder and agoraphobia, taking account of previous methodological problems. One hundred and four patients suffering from DSM III-R panic disorder with or without agoraphobia were randomly allocated to receive cognitive behaviour therapy with either, standard' therapist contact, minimum' therapist contact or as a bibliotherapy. All patients were seen by the same therapist and all received an identical treatment manual. Treatment response, as measured by patient and therapist report scales of anxiety, depression, and agoraphobic avoidance, was analysed in terms of both traditional statistical significance and clinical significance of outcome. At treatment end-point the standard' therapist contact and minimum' therapist contact groups showed significant reductions pre-to posttreatment on all measures. Pre-to post-treatment reductions for the bibliotherapy group were significant on therapist-and patient-rated measures of anxiety only. The standard' therapist contact group was consistently significantly improved in comparison to the bibliotherapy group. Significant differences between the standard' and minimum' therapist contact groups were found on therapist-rated anxiety only. Assessment of clinical significance of treatment outcome showed further differences between treatment groups with the standard' therapist contact group showing the largest proportion of patients achieving clinically significant change on all measures both at treatment end-point and at 6-month follow-up.