BACKGROUND: Some studies support the view that meaningful auditory input can be processed by the brain during apparent surgical anaesthesia. Consequently, patients may be able to remember some information implicitly after anaesthesia as well through a 'dream-like process' (subconscious awareness). The aim of this study was to investigate the presence of subconscious awareness during anaesthesia and to examine its relationship to the mid-latency auditory evoked responses (MLAERs). METHODS: We studied 40 patients, ASA I-II, undergoing laparoscopic cholecystectomy. General anaesthesia was induced with thiopental 5 mg kg(-1), fentanyl 3 micro g kg(-1), and vecuronium 0.08 mg kg(-1). For the maintenance of anaesthesia, patients were randomly assigned to one of four anaesthetic regimen groups: sevoflurane+air in oxygen 40%; sevoflurane+nitrous oxide 60%; isoflurane+air in oxygen 40%; and isoflurane+nitrous oxide 60%. MLAERs were recorded before anaesthesia, at 1 MAC of inhaled anaesthetic and then 30 min after awakening. An audiotape with one of four stories was played immediately after intraoperative MLAER recording. Explicit and implicit memory was assessed 24 h after awakening. RESULTS: None of the patients had explicit recall. One of the patients from the isoflurane-air group showed implicit memory of listening to the audiotape. A dream-like process, in which they remembered implicitly the story played during anaesthesia, occurred in one of the patients from the sevoflurane-nitrous oxide group. In the patients with subconscious awareness, MLAERs were similar to that of the awake state with a Pa latency increase of less than 8.87. When there was a marked increase in Pa latency during anaesthesia, no subconscious awareness was observed. No statistically significant differences were found between Pa latency before and after anaesthesia. CONCLUSIONS: MLAERs may help to predict subconscious cerebral processing of auditory inputs during anaesthesia.