We present a 59-yr-old Accepted for publication 23rd June, 1993. vestigation revealed a dumb-bell shaped cervical meningioma at the C2-3 level of the spinal cord. He had diabetes, well controlled by diet, and mild, untreated hypertension with evidence of left ventricular hypertrophy (LVH) and right bundle branch block (RBBB) on the electrocardiogram (ECG). Dexamethasone and ranitidine were started two days before the cervical laminectomy.No premedication was given. Anaesthesia was induced with thiopentone 125 mg, fentanyl 300 ~g and succinylcholine I00 mg. Following tracheal intubation, anaesthesia was maintained with nitrous oxide 70% and isoflurane 0.5-1% in oxygen, and an infusion of atracurium to maintain the post-tetanic twitch count less than five. t Monitoring consisted of ECG, pulse oximeter, central venous pressure (CVP), direct arterial pressure, oesophageal stethoscope, end-tidal carbon dioxide (PETCOz), and urinary catheter. The central venous catheter (Cavafix 375: B Braun, Melsungen, Germany) was inserted through the right basilic vein and its position at the right atrium was confirmed subsequently by chest x-ray (CXR). Elastic stockings were used to minimize venous pooling. The patient was moved gradually to the sitting position. Normal saline (NS) 500 ml and stable plasma protein solution (SPPS) 500 ml were infused to maintain haemodynamic stability. Three hours after the start of surgery, during dissection of the dura, the PETCO2 decreased from 34.5 mmHg to 28.5 mmHg over 30 sec without any change in arterial pressure, heart rate, minute volume or airway pressure. Bilateral neck compression was performed and the inspired oxygen fraction (FIOz) increased to 1.0. There was no air aspirated from the central line and no site of air entrainment could be found. The PETCO2 gradually returned to normal over five minutes and surgery continued. An FIO2 of 0.3 in nitrous oxide was reintroduced ten minutes later without decrease in oxygen saturation. A similar episode occurred one hour later with a decrease in PETCO2 from 33 mmHg to 21 mmHg over 30 sec but again no source of air entrainment was found. There was no change in arterial pressure, heart rate or oxygen saturation. A third decrease in PETCO2 from 34.5 mmHg to 15 mmHg (Figure 1) developed half an hour CAN J ANAESTH 1993 / 40:10 / pp964-7