BACKGROUND: During one lung ventilation (OLV) for thoracic surgery, hypoxic pulmonary vasoconstriction (HPV) may reduce venous admixture and ameliorate the decrease in arterial oxygenation by diverting blood from the non-ventilated to the ventilated lung. Volatile anesthethics are the drugs of choice in thoracic surgery despite numerous experimental data showing their inhibiting effect on the HPV. The effect of thoracic epidural anesthesia (TEA) on HPV during OLV has not been determined. The object of the present study was to compare the effects of TEA with those of isoflurane on oxygenation and shunting during two lung ventilation (TLV) and OLV in human volunteers. METHODS: Thirty patients who needed OLV for elective thoracic surgery were randomly assigned to receive either group isoflurane (1 MAC isoflurane + saline 6 ml TEA + 0.5 - 1ng/dl propofol + fentanyl + vecuronium, n = 15) and group TEA (0.5% bupivacaine 6 ml TEA + 0.5 - 1ng/dl propofol + fentanyl + vecuronium, n = 15) with 100% oxygen in separate groups. Systemic hemodynamic parameters were recorded, and blood gas values were obtained 30 min after the start of TLV and 30, 45 and 60 min after the start of OLV in the lateral position. RESULTS: Reductions in PaO2 and increases in shunt fraction at all study times after the start of OLV were observed. However, there were no differences in percentage changes between the groups. The other blood gas data (SaO2, SO2, PaCO2, PCO2, pH, Hb, CaO2, CO2) and systemic hemodynamics (mean arterial blood pressure, central venous pressure) did not change at all the study times after the start of OLV in the two groups. CONCLUSIONS: In clinical practice, isoflurane and TEA for OLV was no different in terms of arterial blood oxygenation and pulmonary shunt.