Objectives This randomized controlled trial investigated whether the regional cerebral oxygenation saturation (rScO2)-guided lung-protective ventilation strategy could improve brain oxygen and reduce the incidence of postoperative delirium (POD) in patients older than 65 years.Methods This randomized controlled trial enrolled 120 patients undergoing thoracic surgery who received one-lung ventilation (OLV). Patients were randomly assigned to the lung-protective ventilation group (PV group) or rScO2-oriented lung-protective ventilation group (TPV group). rScO2 was recorded during the surgery, and the occurrence of POD was assessed.Results The incidence of POD 3 days after surgery-the primary outcome-was significantly lower in the TPV group (23.3% versus 8.5%). Meanwhile, the levels of POD-related biological indicators (S100 beta, neuron-specific enolase, tumor necrosis factor-alpha) were lower in the TPV group. Considering the secondary outcomes, both groups exhibited a lower oxygenation index after OLV, whereas partial pressure of carbon dioxide and mean arterial pressure were significantly increased in the TPV group. In addition, minimum rScO2 during surgery and mean rScO2 were higher in the TPV group than in the PV group.Conclusion Continuous intraoperative monitoring of brain tissue oxygenation and active intervention measures guided by cerebral oxygen saturation are critical for improving brain metabolism and reducing the risk of POD.