BACKGROUND: Being awake during anaesthesia is a serious complication. An anaesthetic depth monitor must discriminate in real time between wakefulness and unconsciousness. The present study created a period of wakefulness during propofol-induced hypnosis. Bispectral index (BIS), explicit and implicit memories of the awake period were investigated. METHODS: Ten volunteers were studied. The calculated brain concentration of a target controlled infusion of propofol was increased until loss of response (LOR) to verbal command and then propofol was stopped. When fully awake, volunteers were presented with a picture, sound and smell. Propofol infusion was restarted until LOR and then ceased. BIS and the calculated brain concentration of propofol were recorded every minute. A structured interview was conducted for explicit memories after awakening and for explicit as well as implicit memories the day after. RESULTS: Median BIS-index for the transition between awake and asleep and vice versa differed significantly. It was not possible, however, to establish any threshold value or zone for discriminating between wakefulness and LOR due to the large inter-individual variations in BIS-index. No volunteer could explicitly recall any of the stimuli presented during the period of wakefulness. CONCLUSION: The BIS-index decreases with increasing sedation but because of the large individual variations, the real-time BIS-index for the individual subject cannot reliably discriminate wakefulness from unconsciousness during propofol infusion. Propofol causes such profound amnesia that lack of postoperative recall does not assure that episodes of awareness have not occurred during propofol-induced hypnosis.