Our previous study showed that the plasma microRNA-221-3p level could serve as a biomarker for major depression or mood. This study aimed to further investigate the role of plasma microRNA-221-3p level in postoperative cognitive dysfunction (POCD). Patients undergoing non-cardiac surgery were randomly assigned according to the inclusion and exclusion criteria. POCD was diagnosed by the Z score method. The relative level of plasma microRNA-221-3p was decided by quantitative real-time polymerase chain reaction. Multiple logistic regression analysis and receiver operating characteristic(ROC) curves were used for the analysis of plasma microRNA-221-3p prediction performance for POCD. At 7 days post-surgery, the rate of POCD was 34.04%. Patients in the POCD group had a higher preoperative depression score, older age, and longer operation duration than that in the NPOCD group. The relative level of plasma microRNA-221-3p in the POCD group was 1.78 and 2.73 times higher than that in the NPOCD group at 1 day before and 7 days after the surgery, respectively. The relative content of plasma microRNA-221-3p at 7 days after operation was an independent risk factor for POCD. The ROC curves showed that the area under the curve was 0.938 for plasma microRNA-221-3p at postoperative 7 days, and the threshold for POCD detection was 12.33 with a sensitivity and specificity of 81.3% and 96.3%, respectively. Our results indicate that the plasma postoperative microRNA-221-3p levels could be an effective predictor for POCD after non-cardiac surgery.