This prospective study of patients with acute respiratory illness or potential ventilatory compromise compared pCO(2) and pH on an arterial and a venous blood sample with the aims of determining whether venous pH and pCO(2) can replace arterial values in the management of patients with acute respiratory disease and to determine whether there is a cut-off level of venous pCO(2) that can accurately screen for significant hypercarbia (pCO(2) > 50 mm Hg). Data were analyzed using bias plot and receiver operator characteristic (ROC) curve methods. There were 196 sample-pairs analyzed; 56 (29%) had significant hypercarbia. For pH, there was very good agreement with venous samples being an average of 0.034 pH units lower than arterial samples. With respect to pCO(2), there was only fair agreement, with the pCO(2) on average 5.8 mm Hg higher in venous samples and 95% limits of agreement -8.8 to +20.5 mm Hg. The ROC curve analysis showed that a venous pCO(2) level of 45 mm Hg was a potential screening cutoff (sensitivity for the detection of hypercarbia of 100%, specificity 57%). This study shows that venous pH is an acceptable substitute for arterial measurement but there is not sufficient agreement for venous pCO(2) to be able to replace arterial pCO(2) in the clinical evaluation of ventilatory function. Venous pCO(2) may be able to be used as a screening test for hypercarbia using a screening cut-off of 45 mm Hg.