Objective: To compare the analgesic effect of adductor canal block with different infusion methods for patients undergoing total knee arthroplasty. Methods: A total of 60 elderly patients undergoing total knee arthroplasty, aged 65–77 years, with American Society of Anesthesiologists (ASA) II or III, were randomly divided into two groups (n=30): a programmed intermittent bolus infusion group (Group P) and a constant infusion group (Group C). After the operation, a loading dose of 10 mL of 0.2% ropivacaine was given, and patient-controlled nerve block analgesia pump was connected. Programmed intermittent bolus infusion with a single automatic dose of 10 mL and intermittent time of 2 hours was used in Group P. Constant infusion with a background dose of 5 mL/h was used in Group C. The single patient pressing dose was 5 mL with a 30-minute locking time in both groups. When Visual Analog Scale (VAS) score was over 4 points, the analgesia pump was pressed. If pain was not relieved 30 minutes after pressing by patients, oxycodone hydrochloride of 5 mg was subcutaneously injected as analgesic remedy. VAS and quadriceps strength were recorded 4, 8, 24, 48 and 72 hours after surgery. The walking distance was recorded on day 1–3 after total knee arthroplasty (TKA). The analgesic remedy, patient satisfaction and adverse reactions were recorded within 72 hours after surgery. Results: Compared with Group C, the the dynamic VAS score between 8 hours and 48 hours after surgery was lower in Group P (P<0.05). Compared with Group C, the walking distance on day 1–3 was increased in Group P (P<0.05). But the quadriceps strength, analgesic remedy, patient satisfaction and the incidence of adverse reactions were not significantly different between Group P and Group C (P>0.05). Conclusion: Adductor canal block with programmed intermittent bolus infusion is effective for postoperative analgesia for patients undergoing total knee arthroplasty, and its effect of inhibiting dynamic pain is better than that of constant infusion.