BACKGROUND Measurements of transcutaneous carbon dioxide tension (PtcCO(2)) with current devices are proven to provide clinically acceptable agreement with measurements of partial arterial carbon dioxide tension (PaCO(2)) in several settings but not during cardiopulmonary exercise testing (CPET). OBJECTIVE The primary objective of this study was to investigate the agreement between PaCO(2) and PtcCO(2) measurements (using a Tosca 500 with a Tosca sensor 92) during CPET. A secondary objective was to investigate the agreement between arterial and transcutaneous oxygen saturation (SaO(2), SpO(2)) as measured with this sensor during CPET. METHODS In patients with various pulmonary diseases, PtcCO(2) and SpO(2) were continuously measured and compared with arterial blood gas samples during CPET. A maximum bias of 0.5 kPa and 95% limits of agreement (LOA) of 1 kPa between carbon dioxide pressure (PCO(2)) measurements were determined as clinically acceptable. RESULTS In total 101 'paired' arterial and transcutaneous measurements were obtained from 21 patients. Bias between PaCO(2) and PtcCO(2) was -0.03 kPa with LOA from -0.78 to 0.71 kPa. Bias between SaO(2) and SpO(2) was -1.0% with LOA from -2.83 to 0.83%. CONCLUSIONS Transcutaneous estimations of PCO(2) and SpO(2) are accurate and can be used in CPET, circumvening the need for arterial cannulation.