OBJECTIVE To compare anesthesia with lidocaine 2% by bolus or by continuous perfusion through a subarachnoid catheter. PATIENTS AND METHODS Thirty-one patients undergoing surgery on a lower limb were studied prospectively. Patients were divided into two groups: 14 received subarachnoid anesthesia in continuous perfusion (group CP) while 17 were given subarachnoid anesthesia in bolus form (group B). An 18G Tuohy needle was used with a 20 G catheter. An initial bolus of 20 mg (1 ml) was administered in group CP and followed by continuous perfusion at a rate of 16 mg/h (0.8 ml/h). In group B an initial 20 mg bolus was followed by additional ones of 10 mg (0.5 ml). The optimum sensory level was T 10. RESULTS The two groups were homogeneous with respect to age, weight, height and medical history. Fewer boluses were needed with continuous perfusion. Seven patients (50%) in group CP required additional boluses. The ratio of total drug dose administered in mg to time of surgery was significantly higher in group CP. The sensory and motor blockade levels obtained were similar in the two groups. One case of hypotension and one of bradycardia were recorded. Two cases of minor headache were recorded in group B. Sedation was used in 7 patients (50%) in group CP and in 13 (76%) in group B. CONCLUSION Subarachnoid administration of lidocaine 2% in continuous perfusion affords the same sensory and motor blockade as administration in bolus form and does not reduce the overall amount of lidocaine used.