OBJECTIVE To evaluate the usefulness of transcutaneous carbon dioxide pressure (TcPCO2) monitoring in patients hospitalized for respiratory disease. METHODS We used a SenTec TcPCO2 monitor that also determines transcutaneous oxygen saturation (SpO2) by means of a sensor placed behind the ear lobe at a temperature of 42 degrees C. We compared arterial blood gas measurements--PaCO2 and arterial oxygen saturation (SaO2)--with transcutaneous measurements and analyzed the correlation, regression line, and agreement between the 2 methods. RESULTS Thirty patients (20 men and 10 women) with various respiratory diseases and a mean (SD) age of 71 (13) years were included in the study. The median TcPCO2 was 43.25 mm Hg and the median PaCO2 was 42.6 mm Hg with no significant differences between the 2 measurements. The correlation was significant (rho=0.979; P< .0001) and the corresponding regression equation was TcPCO2=-2.475+1.058 PaCO2. The mean difference was 0.16 mm Hg (95% confidence interval [CI], --0.74 to 1.06). The lower limit of agreement (mean -1.96 SD) was -4.64 mm Hg, and the upper limit (mean +1.96 SD) was 4.96 mm Hg. For SaO2, the median was 94% and for SpO2, 95%. The difference between the 2 medians was significant (P< .004). The correlation was also significant (rho=0.822; P< .0001) with SpO2=4.427+0.97 SaO2. The mean difference was 1.14% (95% CI, 0.381% to 1.899%). The lower limit of agreement (mean -1.96 SD) was --2.93% and the upper limit (mean +1.96 SD) was 5.21% CONCLUSIONS Transcutaneous determination of carbon dioxide pressure and oxygen saturation is useful for patients hospitalized for respiratory disease in view of its good correlation and agreement, although SpO2 does tend to overestimate SaO2.