BACKGROUND: Potent inhalational agents are widely used for thoracic anesthesia. They have several desirable properties, including ease of administration, rapid onset and offset, and bronchodilation. One potential drawback is their ability to directly inhibit hypoxic pulmonary vasoconstriction. Desflurane does not directly inhibit hypoxic pulmonary asoconstriction in vivo, in contrast to isoflurane using the same animal model. In this study, we compared the effects of desflurane and isoflurane on arterial oxygenation and hemodynamics during one lung ventilation in a humans. METHODS: Thirty five patients scheduled for coronary artery bypass graft were randomly assigned to one of group D/I and I/D. Group D/I consisted of four steps. Hemodynamics and oxygenation parameters were checked in each step. Step 1 was checked when they received desflurane to an end tidal concentration of 6% in 93% oxygen from induction until the end of 30 min of two lung ventilation. Step 2 was checked at the end of 30 min after starting one lung ventilation. Step 3 was checked at the end of 30 min after receiving isoflurane to an end tidal concentration of 1.2% in 93% oxygen during one lung ventilation. Step 4 was checked at the end of 30 min after restarting two lung ventilation. Group I/D received the two anesthetic agents in reverse order. We used the simple cross-over design methodology for treatment and period effect. RESULTS: We found no significant difference in hemodynamic and oxygenation parameters between the two inhalational agents except for a significant increase in mean pulmonary arterial pressure and pulmonary capillary wedge pressure caused by desflurane. CONCLUSIONS: During one lung ventilation, the choice between desflurane and isoflurane does not significantly influence arterial oxygenation and shunt fraction but desflurane should be administered with great caution if it is used as an alternative anesthetic in patients with ischemic heart disease.