1. Thoracic gas volume at end expiration (functional residual capacity, FRC) was measured in chronically and acutely hypoxic anaesthetized rats by a plethysmograph method. 2. FRC, measured during air breathing, was 34-62% larger in rats which had been kept in an environmental chamber in 8, 10 or 12% O2 for 3 weeks than in littermate controls. FRC returned to normal after the rats had returned to air for 9 days. There was no constant difference in the pattern of breathing between control and chronically hypoxic rats. 3. Pressure-volume curves measured post mortem showed no difference in the volume of the lung at 25 cm H2O pressure or in the compliance of the lung between chronically hypoxic and control rats. Thus there was no gross mechanical change in the lung to account for the increase in FRC. 4. Acute hypoxia caused by breathing 12% O2 increased FRC in control but not in chronically hypoxic rats. The increase in FRC in control rats was abolished by combined blockade of the vagus nerves and carotid bodies (with procaine) but not by vagal blockade alone. 5. The combined vagal and carotid body blockade reduced FRC significantly in rats which had been in 10% O2 for 3 days but not in those which had been in 10% O2 for 21 days. 6. Lung area measured from radiographs was not reduced by a muscle relaxant in chronically hypoxic rats. Electromyograms from anterior intercostal muscles and the diaphragm showed no electrical activity in expiration in chronically hypoxic rats which might indicate an active muscular basis for their increased FRC. However when FRC was raised by acute hypoxia in control animals there was also no increase in electrical activity in expiration which could have explained their increase in lung volume. 7. We concluded that the increase in FRC during acute hypoxia in control rats was probably due to a reflex from the carotid body. The increase in FRC in chronically hypoxic rats, which was present while they breathed air, may have had an active muscular component in the early stages but later on there was possibly a structural factor in the chest wall.