Introduction:Postoperative neurocognitive disorders (PNDs) are common postoperative complications that can hinder patients' postoperative recovery. Various studies have investigated the correlations between PNDs and cerebrospinal fluid (CSF) neuroinflammatory biomarkers in orthopedic patients. However, combined evidence is required to confirm the homogeneity and robustness of these findings.Methods:Observational studies were searched to explore the associations between PNDs and these biomarkers in orthopedic patients. A comprehensive study retrieval was performed in MEDLINE (via OVID), EMBASE, and the Cochrane Library without any restrictions on language or date. Subgroup and sensitivity analyses were performed to confirm robustness of the results.Result:A total of 27 articles were included in this study. Significant concentration differences were found between PNDs and non-PNDs groups in the majority of preoperative CSF neuroinflammatory biomarkers, particularly in patients with postoperative delirium (POD) and delayed neurocognitive recovery (dNCR). Moderate quality evidence identified that increased preoperative CSF levels of t-Tau (OR: 1.008, 95%CI: 1.005-1.010) and p-Tau (OR: 1.077, 95%CI: 1.042-1.078), along with decreased A beta 42 level (OR: 0.998, 95% CI: 0.997-0.999), were risk factors for POD in orthopedic patients. Low quality evidence suggested that preoperative CSF level of A beta 42/t-Tau was a valuable predictive biomarker for dNCR.Conclusion:Different subtypes of PNDs after orthopedic surgery have different CSF biomarkers, with POD having the most and postoperative neurocognitive disorder (POCD) having the fewest. Studies concerning dNCR and POCD are needed to investigate their correlations with these biomarkers. Meanwhile, studies concerning diagnostic tests are also highly needed to help screen for practical biomarkers and analyze these biomarkers from a more comprehensive perspective.