During high frequency jet ventilation (HFJV), it has been shown that Pa(CO)(2) can be predicted by capnography when the frequency was temporarily reduced to obtain a steady expiratory CO(2) tension (P et(CO)(2)). The influence of the sampling site of expiratory gas in the airway and the driving pressure of the ventilator on the difference between Pa(CO)(2) and P et(CO)(2) was investigated in ten adult patients who underwent general anesthesia. During HFJV (frequency: 100 bpm, inspiratory duty cycle; 30%), its frequency was temporarily reduced to 10 bpm in twelve different condition; i.e., 3 different driving pressures (Pd; 20, 30, and 40 psi) at 4 different sampling sites (d; 0, 2, 5, and 10 cm in the endotracheal tube). Both P et(CO)(2) and Pa(CO)(2) were measured simultaneously, and their difference was evaluated by using Student's t-test. The difference between Pa(CO)(2) and Pmax(CO)(2) (the maximum P(CO)(2) value in the airway during exhalation) was minimal, when Pd was greater than 30 psi, and d was greater than 5 cm. The present study suggests that better prediction of Pa(CO)(2) can be done by P et(CO)(2) during HFJV, when d and Pd were set as large as possible to obtain stable expiratory P(CO)(2) curve.