Whether anesthetized patients register emotionally charged information remains controversial. We tested this possibility using subanesthetic concentrations of propofol or desflurane. Twenty-two volunteers (selected for hypnosis susceptibility) received propofol and desflurane (on separate occasions, and in a random order) at a concentration 1.5-2 times each individual's minimum alveolar anesthetic concentration (MAC)-awake (or equivalent for propofol). We gave vecuronium, intubated the trachea of each volunteer, controlled ventilation, and then presented a neutral (control) drama or a "crisis" drama stating that the oxygen delivery system had failed, assigning crisis and control dramas in a blinded, randomized, and balanced manner. One day later, interviewers blinded to the assigned drama conducted a 2-h structured interview (including hypnosis) to determine whether the contents of the interviews after crisis and control dramas differed. In addition, messages permitting subsequent assessment of learning of matter-of-fact information (Trivial Pursuit-type question task and a behavior task) were presented at the anesthetic concentration just sufficient to prevent response to command in each volunteer. No analyses of the tasks involving matter-of-fact information revealed learning except one which correlated hypnosis susceptibility with behavior task performance. Both propofol and desflurane suppressed memory of the crisis. Consistent with previous findings for isoflurane and nitrous oxide, propofol and desflurane suppressed learning of matter-of-fact information at concentrations just above MAC-awake, except that volunteers' susceptibility to hypnosis correlated with performance of a behavior suggested during anesthesia. Propofol and desflurane suppressed learning of emotionally charged information at anesthetic concentrations 1.5-2 times MAC-awake (less than MAC), a different result from that previously reported for ether.