This study examines, using a randomised control trial, the question of the optimum base for exposure treatment with agoraphobic clients. The results have relevance for community psychiatric nurses and nurse behaviour therapists working with such problems and indeed for other behavioural practitioners. While the results confirm the already extensive evidence for the efficacy of exposure in vivo for agoraphobic clients, the results show that in general there is no difference between outcome of treatment delivered from the patient's own home and the out-patient clinic. There are, however, a small number of clients who will not attend psychiatric out-patient clinics for treatment but who will subsequently do well with home-based treatment. The reasons for the experimental findings and implications for psychiatric nurses are discussed.