Background: Total knee arthroplasty has become a standard for treatment of endstage knee osteoarthritis. Due to intense and complex knee innervation, there is a need to improve the anaesthetic/analgesic approach to such operations. The aim of this randomised clinical trial was to compare the analgesic efficacy of the classical regimen and two of those based on the nerve blocks. Methods: A total of 60 patients was included and subjected to elective total knee arthroplasty under the general balanced anaesthesia. They were randomised to receive postoperatively (1) only morphine 5-10 mg q6h and paracetamol 1 g q6h (MP), (2) femoral nerve block (FNB) or (3) fascia iliac compartment nerve block (FICNB). Nerve blocks were produced by a single administration of 30-40 ml of bupivacaine 0.5%. Pain intensity, duration of neural block and additional consumption of analgesics were recorded postoperatively. Results: There were no demographic differences among the three groups of patients. Pain intensity was significantly lower in the two nerve block groups than in the MP group. The same two groups also demanded significantly less analgesics postoperatively than the MP group. Cardiovascular control was significantly better in the nerve block groups. There were no significant differences between the FNB and FIC-NB groups of patients regarding any of the studied parameters. Conclusion: In comparison with the classical MP analgesia, use of FNB or FICNB after the elective total knee arthroplasty results in lower pain scores, lower systolic blood pressure and less consumption of analgesics in the immediate postoperative course.