Background Controversy still exists regarding whether medial parapatellar approach (MP) or midvastus approach (MV) is preferable in total knee arthroplasty (TKA) up to now.The aim of this meta-analysis was to compare the clinical and radiological outcomes following the MV or MP for TKA.Methods A comprehensive search of unrestricted-language literature of all studies comparing MP with MV was conducted through the electronic literature databases of PubMed,EMBASE,Cochrane Library,CNKI,VIP,and WANFANG.Retrieval time was from the time when databases were built to October 2013.Manual search of relevant trials,reviews,and related articles was also performed.Outcomes of interest included postoperative knee extensor and flexor function,postoperative pain,patella tilt,and complications.Relative risk (RR) and weighted mean differences (WMD) from each trial were pooled using random-effects or fixed-effects model depending on the heterogeneity of the included studies.A subgroup analysis or a sensitivity analysis was conducted to explore the potential source of heterogeneity when necessary.Results Twenty-one randomized controlled trials (RCTs) comprising 1 188 patients (1 450 knees) were eligible.Our results showed that MV was associated with better early postoperative extension (WMD=-1.26,95％ CI-2.36 to-0.16,P=0.02) and flexion (WMD=10.13,95％ CI 5.36 to 14.90,P ＜0.01),less postoperative pain (WMD=-0.21,95％ CI-0.34 to-0.07,P=0.002),and no greater risk for complications than MP.The patella tilt did not differ significantly between the two groups (WMD=-0.70,95％ CI-1.94 to 0.54,P=0.27).Conclusions MV may be a better approach than MP,as it improves postoperative early joint function and decreases oain.Future multi-center randomized controlled studies with large sample sizes are required to verify the current findings.