For extremity surgery, tourniquet is placed routinely. With deflation of the tourniquet, the metabolic product is flushed into the systemic circulation and theoretically poses a potential for toxic reactions. In actual fact, these are rare events with this technique. Vigilant monitoring will detect cardiovascular depression at this time. By the tourniquet application, lactic acidemia, abnormal coagulopathy, hypotention, hyperkalemia, increased PaCO2, and production of noxious oxygen free radicals were reported following the release of the tourniquet. But the serial changes of metabolic derangement, degree of lactic acidemia following the use of the tourniquet were not exactly known. To confirm the safety of the pneumatic tourniquet use for two hours, the serial changes of lactic acid levels, acid-base status, potassium concentration, concentration of respiratory gaaes (arterial and end-tidal CO2,) and also hypotension, dysrhythmias and respiratory pattern following release of the tourniquet were studied. Patients were anesthetized with 1% halothane, 50% nitrous oxide and 50% oxygen. Ventilation was maintained by the ventilator to keep the end-tidal CO2, to 4.0% just before the release, and then respiratory parameters (respiratory rate, tidal volume) were constantly maintained through the study. The data were measured from arterial samples or monitors with the following interval; just before tourniquet apply (BTA), before tourniquet release (BTR), at 1, 3, 5, 15 and 30 minutes after the tourniquet release (ATR 1 m, 3m, 5 m, 15 m 30 m). Data measured before the tourniquet apply were used as control values. All data were analyzed by the paired t-test with control. Changes of mean values of each time in one parameter were analyzed by one-way ANOVA. Correlationships between the parameters and duration of ischemia induced by the tourniquet were analyzed by simple regression. The results of this study were as follows; 1) The arterial concentration of lactic acid was maximally increased at 3 minutes after tourniquet release and not returned to control value until 30 minutes after tourniquet release. 2) End-tidal CO, was reached to maximal values of 5.3% at 5 minutes after release of tourniquet. Accompanying theses changes, spontaneous respiration was recovered from the controlled ventilation in 11 patients out of 13 and fought with mechanical ventilator due to asynchronism of respiratory cycles. 3) Mild metabolic acidosis showing the decreased arterial pH and increased PaCO2, in arterial blood gas analysis was maintained in 30 minutes following the release of tourniquet. 4) There were no significant changes of concentrations of potassium. 5) Three episodes of mild hypotension were observed out of 13 patients, but dysrhythmias and other significant clinical changes not observed through the study. The above results showed the possibility of lactic acidemia and changes of respiratory pattern by increased PaCO2, after release of the tourniquet may occur. More intent monitoring is needed to the patients who have had the metabolic derangement in acid-base balance and increased intracranial pressure in application of tourniquet on limbs.