The incidence of tourniquet pain was evaluated in two groups of patients with 20 each undergoing orthopedic surgery of the lower extremities during epidural anesthesia using plain solution of either 2% lidocaine or 0.5% bupivacaine. The drugs were administered in a randomized fashion. Measurement of the levels of sensory loss to pinprick and incidence of tourniquet pain were made by blind-trust. The maximum analgesia level, time between 1st injection and onset of pain, time between tourniquet inflation and onset of pain were recorded similarly in both groups of patients. The incidence of tourniquet pain was significantly greater in patients given 2% lidocaine (40%) than in patients given 0.5% bupivacaine (10%). The incidence of pain was not related to the time of tourniquet inflation, because patients in the bupivacaine group had a significant longer duration of tourniquet inflation than did patients in the lidocaine group. The incidence of pain was also not related to tachyphylaxis, because 7 of 8 patients who complained tourniquet pain in lidocaine group received less than 3 injections for maintenance of analgesia when tourniquet pain started. In summary, it is apparent that tourniquet pain occurs less frequently when bupivacaine is employed for epidural anesthesia as compared to lidocaine.