Objective: To assess the mid-to-long term clinical outcomes after anterior cruciate ligament (ACL) revision surgery and to analyze their predictors. Methods: The medical records of 235 patients undergoing ACL revision surgery between Jan. 2001 and Dec. 2015 at Department of Sports Medicine, Peking University Third Hospital were reviewed. Data were collected including demographic information, information related to revision surgery (time and cause of graft failure, date of revision surgery, surgical technique, combined injuries and management, etc.), as well as information related to primary ACL reconstruction (time, cause and mechanism of first-time ACL rupture, date of primary ACL reconstruction, surgical technique, combined injuries and management, etc.). Patients were followed up at least 2 years after revision surgery for clinical outcomes [Tegner score, Lysholm score, and International Knee Documentation Committee (IKDC) subjective knee score]. Post-revision surgeries on the involved knee and the contralateral knee joint were also documented. Multivariate regression model was used to analyze the predictors of clinical outcomes after ACL revision surgery. Results: A total of 166 (70.63%) patients were followed up at a mean of (4.44±2.40) years (2.03-14.63 years). Clinical outcomes improved significantly at the last follow-up from pre-operative level, with the Lysholm, Tegner, and IKDC scores improving from 70.51±21.25, 3.39±1.77, 63.78±15.04 to 88.64±14.36, 4.67±1.739, 80.23±13.31 (P<0.05), respectively. Three (1.81%) patients experienced infection while 39 (23.49%) patients underwent surgery after revision surgery during the follow-up. Compared with that those occurred during sports, graft failure that occurred during daily activities or due to surgical technical errors that led to poorer clinical outcomes, with the Lysholm, Tegner, and IKDC scores of 9.90 (95%CI: 1.49-18.31), 1.41 (95%CI: 0.10-2.72), 10.35 (95%CI: 0.17-20.54), and 8.53 (95%CI: 1.31-15.75), 1.28 (95%CI: 0.14-2.43), 9.39 (95%CI: 1.03-17.74) lower, respectively. Compared with antero-medial portal, transtibial technique for placement of the femoral bone tunnel showed poorer Lysholm scores of 11.18 (95%CI: 4.73-17.63, P=0.001). Concurrent repair of medial meniscus yielded higher IKDC scores of 11.06 (95%CI: 1.21-20.92, P=0.029) than those with intact medical meniscus. Other factors showed no significant effect. Conclusion: ACL revision surgery is able to restore knee stability and improve knee function. Graft failure caused by sports, concurrent repair of medical meniscus and antero-medial portal technique predicts better outcomes after revision surgery.