Purpose: Perioperative neurocognitive disorder (PND) is common in elderly surgical patients and severely affects postoperative recovery. However, effective prevention is still lacking. Potential perioperative cerebral stressors (including inappropriate sedative/ analgesic depth and imbalanced cerebral oxygen supply/demand) may be important contributing factors. We developed an anesthesia management protocol based on multimodal brain monitoring to achieve standardized, individualized, and real-time regulation of sedative/analgesic depth and cerebral oxygen saturation and investigated whether it could reduce the incidence of PND and its underlying mechanisms. Patients and Methods: Patients (aged >= 65 years) were randomized into Groups C (n=88) and E (n=93). Patients in Group E received multimodal brain monitoring-guided anesthesia management, and those in Group C received BIS-guided anesthesia management. The Montreal Cognitive Assessment (MoCA) was performed both before and seven days after surgery. The postoperative pain scores were recorded. Resting-state functional MRI data were analyzed to examine functional connectivity (FC). Results: Group E demonstrated a numerically lower incidence of PND (15.50% vs 21.59% in Group C), but this difference was not statistically significant. Patients in Group E had increased FC within the right pulvinar, right sub-gyral region, and right inferior parietal lobule (P < 0.05). Significantly lower pain scores were observed in Group E at rest (1h: P=0.04; 24h: P=0.04) and during movement (1h: P=0.03). Conclusion: These results suggest that multimodal brain monitoring-guided anesthesia management may protect neurocognition by enhancing FC within cognition-associated brain regions and attenuating postoperative acute pain. And multimodal brain monitoring-guided anesthesia management may confer a clinically relevant reduction in PND incidence compared to BIS-guided management in elderly surgical patients.