In an attempt to establish whether prolonged withdrawal symptoms after stopping intake of benzodiazepines is caused by return of anxiety, hysteria, abnormal illness behaviour or the dependence process itself producing perhaps a prolonged neurotransmitter imbalance, a group of such patients suffering prolonged withdrawal symptoms (PWS) was compared on a range of psychophysiological measures with matched groups of anxious and conversion hysteria patients and normal controls. It was found that the psychophysiological markers of anxiety were not marked in the PWS group; nor were the averaged evoked response abnormalities found to be associated with cases of hysterical conversion in evidence. The PWS group were hard to distinguish from normal controls on the basis of psychophysiological measures and thus it was felt to be unlikely to be an affective disturbance. It was concluded that PWS is likely to be a genuine iatrogenic condition, a complication of long-term benzodiazepine treatment.