The present study of the effect of two common anaesthetic gases on blood acid-base parameters shows that the micro-Astrup measurement of carbon dioxide tension is not invalidated by the presence of nitrous oxide. This result was anticipated from the theoretical aspect of this technique. The mean error involved in estimating plasma carbon dioxide content in the presence of nitrous oxide using the volumetric Van Slyke apparatus without absorption of carbon dioxide by sodium hydroxide can be of the order of 25%. No such effect was measurable in estimating carbon dioxide contents in the presence of halothane. The degree of respiratory alkalosis during anaesthesia reported in earlier papers (Walker, Morgan, Breckenridge, Watt, Ogilvie, and Douglas, 1963; Morgan, Ogilvie, and Walker, 1963) was greater than had been originally appreciated.A ;false' increase in carbon dioxide content will also falsely increase buffer base or ;base excess' as calculated from standard nomograms (Singer and Hastings, 1948; Davenport, 1958; Siggaard-Andersen, 1963).