Background: Tranexamic acid (TXA) was conducive in total knee arthroplasty (TKA) to reduce blood loss and transfusion demand. The purpose of this meta-analysis was to assess the efficacy and safety of different administration of TXA in primary TKA. Methods: Database PubMed, Medline, Web of Science and Embase were searched. The relative risks (RRs) with 95% confidence intervals (CIs) were calculated to analysis dichotomous outcomes. The mean differences (MD) with 95% CIs were calculated to analysis dichotomous outcomes. Data was analyzed using RevMan 5.3. Results: Twenty-eight randomized controlled trials (RCTs) studies were included in this meta-analysis involving a total of 4,200 participants. There were no obvious differences between oral, intravenous or topical TXA group in total blood loss (intravenous vs. topical: MD = 11.55, 95% CI, -10.23 to 33.34, oral vs. intravenous or topical: MD = -52.25, 95% CI, -121.28 to 16.78), transfusion rate (intravenous vs. topical: RR = 1.04, 95%CI, 0.64 to 1.69, oral vs. intravenous or topical: RR =0.75, 95% CI, 0.36 to 1.54), incidence of venous thrombotic events (VTE) (intravenous vs. topical: RR = 1.43, 95% CI, 0.81 to 2.54). The topical TXA administration had significantly increased postoperative hemoglobin (HB) level compared with the intravenous TXA administration (MD = -0.37, 95% CIs, -0.47 to -0.26). In the combined group, the total blood loss (MD = -119.58, 95% CI, -181.68 to -57.49) and postoperative HB level (MD = 0.54, 95% CI, 0.45 to 0.64) were more acceptable than the single-route group. Conclusions: Combined administration of TXA can reduce total blood loss, postoperative HB drop compared with intravenous, topical or oral TXA alone. Oral administration of TXA is similar to intravenous or topical TXA use alone.