The rebreathing method of measuring oxygenated mixed venous Pco(2) (Pvco(2)) was originally introduced as a bloodless way to estimate arterial Pco(2) (Paco(2)). It has become common practice to subtract 6 mm Hg from the Pvco(2) to obtain the Paco(2) but there are many circumstances in which this leads to an overestimate of the Paco(2). Measurements of Pvco(2) and Paco(2) in 19 patients have shown that a better approximation to Paco(2) under normal conditions of cardiac output and arterial O(2) saturation is Paco(2) = 0.8 Pvco(2). These studies also showed that the Pvco(2) - Paco(2) difference may be much wider, particularly in the presence of arterial unsaturation and a low cardiac output.The factors governing the venoarterial Pco(2) difference are reviewed and their magnitude is calculated to emphasize the complementary roles of measurements of Pvco(2) and Paco(2) in the assessment of patients with cardiorespiratory disease.