Patients undergoing bronchoscopy are usually monitored with pulse oximetry to measure arterial oxygen saturation, but this can fail to detect hypoventilation, particularly if added inspired oxygen is used. Transcutaneous oxygen and carbon dioxide tensions can be measured; the later reflecting respiratory drive. We compared transcutaneous PO2 and PCO2 values with oxygen saturation in patients undergoing day-case bronchoscopy, to see if this information would further improve the safety of the bronchoscopic procedure. Twenty-two consecutive patients undergoing routine fibreoptic bronchoscopy (15 male, mean age 62.3 years; range 45-82 years), were studied using pulse oximetry (OXImeter, Radiometer) and transcutaneous PCO2/PO2 monitoring (TCM3, TINA, Radiometer). We documented a statistically significant increase in transcutaneous PCO2 from mean (SD) stable baseline levels of 5.8 (0.3) kPa (range 4.2-7.9 kPa) to mean peak levels during bronchoscopy of 7.0 (1.0) kPa (range 5.0-8.7 kPa). The time to first adverse change in transcutaneous PCO2 (P = 0.046) and PO2 (P = 0.035) occurred more rapidly than reduction in oxygen saturation in 19 of the 22 cases; median times for change in PCO2 of 67 s (range 10-1800 s), PO2 of 120 s (range 26-559 s) and oxygen saturation of 174 s (range 43-1332 s), timed from administration of i.v. sedation prior to each bronchoscopy. Transcutaneous PCO2/PO2 monitoring during fibreoptic bronchoscopy provided evidence of hypoventilation with significantly elevated levels of transcutaneous PCO2. This method of monitoring provides an earlier indication of respiratory depression during fibreoptic bronchoscopy compared with pulse oximetry.