Eighteen patients with chronic specific phobias were allocated at random to exposure for two hours to the real phobic situation (flooding) under one of three conditions: (1) exposure starting four hours after oral diazepam 0.1 mg./kg. (`waning' group); (2) exposure starting one hour after oral diazepam 0.1 mg./kg. (`peak' group); (3) exposure starting four hours or one hour after oral placebo. Each patient had two treatment sessions in a balanced sequence, and each patient had either one or two of the three treatment conditions. Assessment was blind for the patient, therapist and independent rater.
Exposure under all three conditions produced significant improvement. However, the `waning' group was significantly superior to placebo, and the `peak' group was in between. This trend of superiority was consistent for clinical, attitudinal and physiological ratings. The trend was already present significantly after the first treatment session, before crossover. The addition of diazepam did not increase treatment pleasantness. There was no evidence of drug dissociation at the moderate dosage employed. Serum levels of diazepam varied widely between patients, but were reliable within patients over two sessions. Serum levels bore no relation to psychotropic effect of diazepam.