Apneic, anesthetized patients frequently develop airway obstruction or may be disconnected from ventilatory support. The rate of PaCO2 rise is usually assumed to be equal to that of anesthetized humans who are receiving apneic oxygenation. Apneic oxygenation may eliminate CO2 because it requires a continuous O2 flow. The CO2 rate of rise in anesthetized humans with airway obstruction was measured. Fourteen consenting healthy adults were monitored continuously with pulse oximetry and EKG. Enflurane--O2 anesthesia was established for at least 10 minutes with normal PaCO2 without neuromuscular blockade so that anesthesia was deep enough to prevent spontaneous ventilation. Then, patients' tracheal tubes were clamped. Arterial blood samples were obtained before and after 0, 20, 40, 60, 120, 180, 240, and 300 seconds after clamping, provided that oxyhemoglobin saturation exceeded 0.92. The equation that best described the PaCO2 rise was a logarithmic function. Piecewise linear approximation yielded a PaCO2 increase of 12 mmHg during the first minute of apnea, and 3.4 mmHg/minute thereafter. These values should be employed when estimating the duration of apnea from PaCO2 change for anesthetized patients who lack ventilatory support. In addition, it appears that the flows of O2 that most earlier investigators used when delivering apneic oxygenation probably did not eliminate significant CO2 quantities.