Background Total knee arthroplasty (TKA) is a surgical procedure that often demands potent analgesia. Anesthetists commonly rely on continuous epidural analgesia or psoas compartment block. After the American Society of Regional Anesthesia (ASRA) recent recommendation against the implication of deep regional blocks in the setting of perioperative anticoagulation, femoral nerve block might be a suitable alternative technique. Its effectiveness in TKA is a questionable issue. Our study is hypothesizing that continuous femoral block could provide an analgesic profile similar to continuous psoas compartment block. Patients and method Sixty patients undergoing TKA were randomly assigned into two groups to receive continuous femoral block or continuous psoas compartment block using ultrasound and nerve stimulation guidance. A bolus of 20 ml 0.25% levobupivacaine was injected through a perineural catheter in each group before surgery followed by a 5 ml bolus of 0.125%, and then 5 ml/h infusion at completion of surgery. A bolus of 5 ml 0.125% levobupivacaine was injected when numerical rating scale (NRS) was at least 5. In the first 24 h postoperatively, the median value and the interquartile range of the NRS at 1, 6, 12, and 24 h and after the first physiotherapy session were recorded as the primary end point. The total levobupivacaine and morphine consumption during the same 24-h period were the secondary end points, apart from patient's satisfaction with analgesia during hospitalization. Results The median and interquartile range of the NRS were 4.00 and 2.00 in the first hour, 4.00 and 1.5 in the sixth hour, 4.00 and 1.5 in the 12 th hour, 4.00 and 2.00 in the 24 th hour, and 4.00 and 2.00 after the first physiotherapy session, respectively, in the Fm group, and 3.00 and 3.00 in the first hour, 4.00 and 2.00 in the sixth hour, 4.00 and 1.5 in the 12 th hour, 3.00 and 2.5 in the 24 th hour, and 3.00 and 2.00 after the first physiotherapy session, respectively, in the Ps group. The total mean levobupivacaine and morphine consumption in the Fm and the Ps groups were 218. ± 7.89 vs. 216.6 ± 8.42 and 15.5 ± 4.01 vs. 13.50 ± 4.38, respectively, with no statistically significant difference between both the groups. The catheter insertion time was significantly shorter in the Fm group (8.53 ± 3.1 min) compared with (10.5 ± 3.2 min) the Ps group ( P = 0.021). Conclusion Continuous femoral nerve block provides a suitable alternative analgesic technique in patients undergoing TKA when psoas compartment block is unwarranted in the setting of perioperative anticoagulation as recommended by ASRA.