Purpose: To compare the effects of isoflurane and sevoflurane on arterial oxygenation and middle cerebral artery blood flow velocity during one lung ventilation. Methods: This was a randomized, crossover study in 20 patients undergoing thoracotomy for oesophageal cancer and scheduled for long term one lung ventilation (OLV). They were randomized to one of two groups: group A, firstly isoflurane was administered followed by sevoflurane, and then isoflurane was resumed; group 13, the order of the administration was reversed, Arterial blood gas samples were drawn at the start of OLV, 30 and 60 min after the initiation of OLV and the end of OLV (the change of volatile anesthetics was done 30 and 60 min after the start of OLV). Middle cerebral artery (MCA) was monitored continuously with the probe positioned over the temporal bone window. This probe transmitted 2 MHZ wave Doppler signals. Time-averaged MCA blood flow velocity was calculated from the signals. Results: The PaO2 values decreased 30 min after the start of OLV (364.4 +_ 33.4 mmHg vs 179.0 +_ 19.5. and 338,7 _+ 24.8 mmHg vs 139.7 ___ 19.9 in groups A and 13 respectively), but there was no difference between the groups, Blood flow velocity of MCA did not change after the start of OLV (53. I ___ 3.2, 55.9 _+ 3.0, 56.4 ___ 2,4, and 54. I _+_ 1.9 vs 50.8 _+ 2, I, 50.7 +--2.4, 53.7 • 1.5, 50.8 -+ 2.2 cm'sec-' in groups A and 13 respectively): there was no difference between the groups. (P < 0.05). Conclusion: In clinical practice, the selection of either isoflurane and sevoflurane for OLV was of no difference in terms of the artenal blood oxygenation. With both agents MCA blood flow velocity was maintained dudng OLV.