We designed this study to evaluate the effect of small-dose IV ketamine in combination with continuous femoral nerve block on postoperative pain and rehabilitation after total knee arthroplasty. Continuous femoral nerve block was started with 0.3 mL/kg of 0.75% ropivacaine before surgery and continued in the surgical ward for 48 h with 0.2% ropivacaine at a rate of 0.1 mL . kg(-1) . h(-1). Patients were randomly assigned to receive an initial bolus of 0.5 mg/kg ketamine followed by a continuous infusion of 3 mug . kg(-1) . min(-1) during surgery and 1.5 mug . kg(-1) . min(-1) for 48 h (ketamine group) or an equal volume of saline (control group). Additional postoperative analgesia was provided by patient-controlled IV morphine. Pain scores and morphine consumption were recorded over 48 h. The maximal degree of active knee flexion tolerated was recorded daily until hospital discharge. Follow-up was performed 6 wk and 3 mo after surgery. The ketamine group required significantly less morphine than the control group (45 +/- 20 mg versus 69 +/- 30 mg; P < 0.02). Patients in the ketamine group reached 90 degrees of active knee flexion more rapidly than those in the control group (at 7 [5-11] versus 12 [8-45] days, median [25%-75% interquartile range]; P < 0.03). Outcomes at 6 wk and 3 mo were similar in each group. These results confirm that ketamine is a useful analgesic adjuvant in perioperative multimodal analgesia with a positive impact on early knee mobilization. No patient in either group reported sedation, hallucinations, nightmares, or diplopia, and no differences were noted in the incidence of nausea and vomiting between the two groups.