BACKGROUND: Both postoperative pain control and range of motion are important in total knee arthroplasty (TKA). However, in the literature, there is little comparison of peripheral nerve blocks and periarticular infiltration techniques using levobupivacaine. The aim of our study was to measure pain with visual analogue scale (VAS) and knee range of motion (ROM) between in patients undergoing adductor canal block (ACB) for TKA using levobupivacaine compared to periarticular levobupivacaine infiltration (PAI-L). DESIGN AND SETTING: Prospective randomized clinical trial in a university hospital. METHODS: Patients aged 40-85 years who underwent unilateral TKA were included; 39 were treated with periarticular infiltration using 40 ml (0.125 mg) of levobupivacaine (PAI-L group); and 40 were treated with ACB using 20 ml of 0.25% levobupivacaine (ACB-L group). Postoperative pain scores at rest and during active physical therapy were assessed using a VAS, along with knee ROM in flexion and extension. In addition, 100-foot walking time results, total morphine consumption and time of first analgesia requirement were recorded postoperatively. RESULTS: VAS scores at rest and during active physical therapy and the total amount of morphine consumed were lower in the ACB-L group than in the PAI-L group (P < 0.05). In contrast, knee ROM in flexion and extension and 100-foot walking times were greater in the PAI-L group than in the ACB-L group (P < 0.05). CONCLUSION: ACB-L was superior to PAI-L regarding pain treatment after TKA; however, PAI-L was superior to ACB-L regarding postoperative ROM and walking ability. CLINICAL TRIAL REGISTRY: ACTRN-12618000438257.