The slow response characteristics of the combined transcutaneous electrode have been viewed as a major disadvantage when compared with other types of non-invasive assessment of gas exchange during exercise testing. We have previously shown that by using the highest recommended temperature of 45 degrees C to reduce response times, and combining this with an exercise protocol of gradual work load increments, that this allows changes in arterial blood gases to be closely followed by transcutaneous values. In the present study we have validated the use of a transcutaneous electrode for estimation of alveolar-arterial oxygen gradient (AaO2) and dead space to tidal volume ratio (V(D)/V(T)) during exercise, against values calculated from direct arterial blood gas analysis. One hundred measurements were made in 20 patients with various cardiopulmonary disorders who underwent exercise testing. Exercise testing was performed by bicycle ergometry with a specific protocol involving gradual work load increments at 2 min intervals. Transcutaneous gas tensions were measured by a heated combined O2 and CO2 electrode. Arterial blood was sampled at the midpoint of each stage of exercise and transcutaneous tensions noted at the end of each stage. The mean difference of the AaO2 gradient calculated from blood gas tensions obtained by the two methods was 0.14 kPa. The limits of agreement were -0.26 and 0.63 kPa. The same values for V(D)/V(T) calculated from gas tensions measured by the two methods were: mean difference 0001; limits of agreement -0.0242 and 0.0252. For both these parameters there was an even scatter around the mean value on Bland and Altman analysis. The findings of this study suggest that estimation of parameters of gas exchange using transcutaneous values during exercise testing is reliable, provided the electrode is heated to a slightly higher temperature than usual and the work load increments are gradual, allowing for the latency in the response time of the system. This system allows the assessment of the contribution of ventilation/perfusion inequality to breathlessness on exertion in patients, provided an initial arterial or ear lobe capillary sample is obtained for calibration purposes. This technique is particularly valuable in patients undergoing repeat exercise tests as it circumvents the need for arterial cannulation.