OBJECTIVE: To estimate the early effectivenss of computer navigation-assisted total knee arthroplasty (TKA) by comparing with traditional TKA. METHODS: The clinical data of 89 patients (100 knees) underwent primary TKA between October 2017 and July 2018 were analyzed retrospectively, including 44 patients (50 knees) who completed the TKA under the computer-assisted navigation system as the navigation group and 45 patients (50 knees) treated with traditional TKA as the control group. There was no significant difference between the two groups ( >0.05) in gender, age, body mass index, diagnosis, side, disease duration, Kellgren-Lawrence classification of osteoarthritis, and preoperative American Hospital for Special Surgery (HSS) score, range of motion (ROM), hip-knee-ankle angle (HKA) deviation. The operation time, incision length, difference in hemoglobin before and after operation, postoperative hospital stay, and the complications were recorded and compared between the two groups. The HSS score, ROM, and joint forgetting score (FJS-12) were used to evaluate knee joint function in all patients. Unilateral patients also underwent postoperative time of up and go test and short physical performance battery (SPPB) test. At 1 day after operation, the HKA, mechanical lateral distal femoral angle (mLDFA), mechanical medial proximal tibial angle (mMPTA), sagittal femoral component angle (sFCA), and sagittal tibial component angle (sTCA) were measured and calculated the difference between the above index and the target value (deviation); and the joint line convergence angle (JLCA) was also measured. RESULTS: The operations of the two groups were successfully completed, and the incisions healed by first intention. The operation time and incision length of the navigation group were longer than those of the control group ( <0.05); the difference in difference of hemoglobin before and after the operation and the postoperative hospital stay between groups was not significant ( >0.05). Patients in the two groups were followed up 27-40 months, with an average of 33.6 months. Posterior tibial vein thrombosis occurred in 1 case in each of the two groups, and 1 case in the control group experienced repeated knee joint swelling. The HSS scores of the two groups gradually increased after operation ( <0.05); HSS scores in the navigation group at 1 and 2 years after operation, and knee ROM and FJS-12 scores at 2 years were significantly higher than those in the control group ( <0.05). There was no significant difference in the postoperative time of up and go test and SPPB results between the two groups at 7 days after operation ( >0.05); the postoperative time of up and go test of the navigation group was shorter than that of the control group at 2 years ( =-2.226, =0.029), but there was no significant difference in SPPB ( =0.429, =0.669). X-ray film measurement at 1 day after operation showed that the deviation of HKA after TKA in the navigation group was smaller than that of the control group ( =-7.392, =0.000); among them, the HKA deviations of 50 knees (100%) in the navigation group and 36 knees (72%) in the control group were less than 3°, showing significant difference between the two groups ( =16.279, =0.000). The JLCA and the deviations of mLDFA, mMPTA, sFCA, and sTCA in the navigation group were smaller than those in the control group ( <0.05). CONCLUSION: Compared with traditional TKA, computer navigation-assisted TKA can obtain more accurate prosthesis implantation position and lower limb force line and better early effectiveness. But there is a certain learning curve, and the operation time and incision length would be extended in the early stage of technology application.