The aim of our retrospective study was to evaluate the efficacy of routine pulse oximetry and capnometry for detection of oesophageal tube misplacement. Patients undergoing ENT interventions at our hospital are routinely monitored by ECG, arterial blood pressure by cuff, capnography, and pulse oximetry. Beat-to-beat values of Sao2 and CO2 waveform were recorded by a graphic printer connected to a microcomputer, ASA I patients were routinely preventilated with FIO2 = 0.3, and ASA II-III patients with FIO2 = 1.0. Anaesthesia was performed by junior anaesthesiologists under the close supervision of a resident. During a 16-month period, 1372 patients were anaesthetized. The records of 21 patients with accidental oesophageal tube misplacement were available for retrospective evaluation. Nine patients were preventilated with FIO2 = 0.3 (ASA I), 12 patients with FIO2 = 1.0 (ASA II-III). Rapid detection of oesophageal tube position as early as the first ventilation is possible by capnometry, because of the highly significant difference in end-tidal CO2 (0.2 +/- 0.2 vol%; tracheal intubation: 3.7 +/- 0.9 vol.%; P less than 0.0001). The present advanced pulse oximetry method does not permit differentiation between oesophageal and tracheal tube position within 30 s in patients preventilated with FIO2 = 1.0. Oesophageal misplacement was detectable within 7.5 +/- 0.9 s in patients preventilated with FIO2 = 0.3 due to a 2.1 +/- 0.8% decrease in Sao2 (P less than 0.001). Our results underscore the significance of capnometry for rapid detection of inadvertent oesophageal intubation. High-resolution pulse oximetry is a valuable supplement but not a substitute for capnometry.