During the prospective non-randomized study 340 patients were operated due to bone tumors in the volume of tumor removal with consequent defect replacement with megaprothesis of the hip (n = 52), knee (n = 254), ankle (n = 11) joints and total endoprosthesis of femur (n = 23). The combination of spinal or superficial general anesthesia with epidural (n = 101) or conduction (n = 239) analgesia was used for pain relief. The regional pain relief after surgery continued for 3-4 days. The level of pain, need for additional pain relief, frequency of failures and complications were studied as per the digit rating scale. The following nonparametric analysis methods were used for statistical processing: U-Mann-Whitney test or Fisher’s exact test. The number of patients with acceptable pain level after surgery and the need for additional pain relief were non statistically different in the groups with epidural or conduction pain relief. With post-surgery epidural pain relief compared to conduction analgesia arterial hypotension (12% versus 5%, p = 0.035) and dysuric disorders (5% versus 0.4%, p = 0.010) developed more often. The conduction analgesia turned out to be more complicated from technical point of view. Conclusion: given the similar efficiency of pain relief the conduction analgesia compared to epidural one has been followed by the less number of the side effects.