1. Data have been combined from three previous series to provide revised standards for the prediction of physiological dead-space volume ( V D ), arterial oxygen tension ( P a,o 2 ), alveolar-to-arterial oxygen-tension difference ( P a,o 2 - P a,o 2 ) and venous admixture fraction (Q̇va/Q̇t) in the sitting position. 2. These standards, based on measurements in 96 healthy men and women aged from 20 to 74 years, largely confirm conclusions drawn from the first series of 48 subjects. 3. V D is best predicted on age, height, tidal volume and the reciprocal of respiratory frequency. P a,o 2 , ( P a,o 2 - P a,o 2 ) and Q̇va/Q̇t are adequately predicted on age alone.

1. The alveolar-arterial oxygen-tension gradient was measured in duplicate, in forty-eight healthy subjects (twenty-four men and twenty-four women) aged from 20 to 74 years, while breathing oxygen concentrations of approximately 14, 21, 40, 60 and 100%. 2. The gradient increased with age and with inspired O 2 concentration up to 60%. Above 60%, during normal breathing, there was no significant change in gradient. 3. Breathing 100% O 2 in deep breaths decreased the gradient in all age-groups, significantly so in subjects of 50 years and younger. The remaining gradient was due to anatomical venous shunt. 4. When this anatomical shunt was allowed for, the differences in gradient between age-groups were largely abolished at all inspired O 2 concentrations. The gradient which was not due to anatomical shunt, when breathing air, remained higher in subjects over 60 years than in the younger subjects. 5. The results can be interpreted to mean that three components contribute to the normal gradient: (a) irreversible anatomical shunt; (b) closed, unventilated alveoli which can only be inflated by deep breaths; and (c) cyclical airway closure during normal breathing in subjects over 60 years. Diffusion disequilibrium may contribute a gradient of up to 3 mmHg during the breathing of low O 2 concentrations.