Measurement of end-tidal carbon dioxide (PETCO2) is useful because of its noninvasiveness, continuity, and response time when sudden changes in ventilation occur during inhalation sedation. We compared the accuracy of PETCO2 using a nasal mask and nasal cannula with the accuracy of transcutaneous carbon dioxide (TC-CO2) and determined which method is more useful during inhalation sedation in volunteers. We used a modified nasal mask (MNM) and modified nasal cannula (MNC) for measurement of PETCO2. The capnometer measured PETCO2 in the gas expired from the nasal cavity by means of two devices. The volunteers received supplemental O-2 by means of each device at a flow rate of 6 L/min. After the volunteers lay quietly for 5 min with a supply of 100 % O-2, they received supplemental N2O by means of each device at concentrations of 10, 20, and 25 % for 5 min and 30 % for 25 min. The correlation coefficient was poorer in the MNM than in the MNC, and the mean difference between TC-CO2 and PETCO2 in the MNM was greater than that in the MNC. The difference between the TC-CO2 and PETCO2 ranged from 3 to 6 mmHg in the MNM and from 2 to 5 mmHg in the MNC. The difference between two variables against the TC-CO2 and the CO2 waveforms obtained by means of the two devices were within the clinically acceptable range. Our two devices can provide continuous monitoring of PETCO2 with a supply of N2O/O-2 in patients undergoing inhalation sedation.