Background: Postoperative pain control after total knee arthroplasty (TKA) remains a great challenge. The management of pain in the immediate postoperative period is one of the most critical aspects to allow speedier rehabilitation and reduce the risk of postoperative complications. Recently, periarticular infiltration anesthesia has become popular, but the outcome is controversial. Some studies have shown transient effects, rebound pain, or no effectiveness in pain control. Continuous intra-articular infusion technique has been introduced to improve these transient effects, but more clinical studies are needed. Furthermore, the potential risk of early periprosthetic joint infection is causing concerning. We plan to compare continuous intra-articular infusion anesthesia with epidural infusion anesthesia after TKA to assess the effectiveness of this technique in reducing pain, in improving postoperative function, and to look at the evidence for risk of early infection. Methods/design: This trial is a randomized, controlled study. Patients (n = 214) will be randomized into two groups: to receive continuous intra-articular infusion anesthesia (group C); and epidural infusion anesthesia (group E). For the first 3 postoperative days, pain at rest, active range of motion (A-ROM), rescue analgesia and side effects will be recorded. At 3-month and 6-month follow-up, A-ROM, C-reactive protein, erythrocyte sedimentation rate, and synovial fluid cell count and culture will be analyzed. Discussion: The results from this study will provide clinical evidence on the efficacy of a continuous intra-articular infusion technique in reducing pain, postoperative functional improvement and safety. It will be the first randomized controlled trial to investigate infection risk with local anesthesia after TKA.