Femoral nerve block (FNB), a commonly used postoperative pain treatment after total knee arthroplasty (TKA), reduces quadriceps muscle strength essential for mobilization. In contrast, adductor canal block (ACB) is predominately a sensory nerve block. We hypothesized that ACB preserves quadriceps muscle strength as compared with FNB (primary end point) in patients after TKA. Secondary end points were effects on morphine consumption, pain, adductor muscle strength, morphine-related complications, and mobilization ability. We performed a double-blind, randomized, controlled study of patients scheduled for TKA with spinal anesthesia. The patients were randomized to receive either a continuous ACB or an FNB via a catheter (30-mL 0.5% ropivacaine given initially, followed by a continuous infusion of 0.2% ropivacaine, 8 mL/h for 24 hours). Muscle strength was assessed with a handheld dynamometer, and we used the percentile change from baseline for comparisons. The trial was registered at clinicaltrials.gov (Identifier: NCT01470391). We enrolled 54 patients, of which 48 were analyzed. Quadriceps strength as a percentage of baseline was significantly higher in the ACB group compared with the FNB group: (median [range]) 52% [31-71] versus 18% [4-48], (95% confidence interval, 8-41; P = 0.004). There was no difference between the groups regarding morphine consumption (P = 0.94), pain at rest (P = 0.21), pain during flexion of the knee (P = 0.16), or adductor muscle strength (P = 0.39); neither was there a difference in morphine-related adverse effects or mobilization ability (P > 0.05). Adductor canal block preserved quadriceps muscle strength better than FNB, without a significant difference in postoperative pain.