Background: The efficacy of morphine added to periarticular multimodal drug injection (PMDI) for pain management after total knee arthroplasty (TKA) is controversial. Adding morphine to spinal anesthesia has reportedly improved pain relief for the first 24 h. We examined the effect of morphine added to PMDI or spinal anesthesia on pain management and functional recovery after TKA. Methods: A total of 97 patients were randomized into three groups: in Group A (34 patients), 10 mg morphine was added to PMDI; Group B (31 patients), 0.1 mg morphine was added to spinal anesthesia; and Group C (32 patients), morphine was added to neither the PMDI nor spinal anesthetic. To evaluate the efficacy of added morphine for pain management, we assessed rest pain, the number of times analgesics were used, and the time period until the first analgesic use. The adverse effects of morphine were assessed by counting the numbers of times vomiting occurred and antiemetics were used. Functional recovery was evaluated by recording the range of motion of the knee and the date of ability to walk. Results: Rest pain was the least in Group B at 6 and 12 h after operation. The number of times analgesics were used was the least in Group B. The time period until the first analgesic use was the longest in Group B. The number of vomiting episodes was the least in Group C. The number of times antiemetics were used was higher in Group A than in Group C. There were no significant differences in the range of motion and date of ability to walk among the three groups. Conclusions: The efficacy of morphine added to PMDI was limited, and that of morphine added to spinal anesthesia disappeared within 20 h postoperatively. Adding morphine to PMDI or spinal anesthesia did not improve functional recovery and caused some adverse effects.