‘Dark’ respiratory losses of CO2 were measured on a one year old sward of S24 perennial ryegrass (Lolium perenne L.) at intervals during a 74 day reproductive growth period, between April and June, and a 21 day vegetative growth period, in July and August. Part of the sward was shaded for one week before measure ments commenced. Measurements of ‘dark’ respiration continued for 46 hand it was possible to distinguish two components which are designated ‘maintenance’ and ‘synthetic’ ‘Maintenance’ respiration was taken to be the mean rate of CO2 efflux after 40–46 h darkness. When calculated on a plant d. wt basis at 15°C it ranged between 6 to 32mg CO2 g-1 day-1 during reproductive growth and 10–14 mg CO2 g-1 day-1 during vegetative growth. During reproductive growth, sward protein content ranged between 7–23 per cent and when maintenance respiration was recalculated on the basis of protein content it changed relatively little throughout the growth period (90–140 mg CO2 g pro tein-1 day-1); the value for vegetative growth ranged between 70–100mg CO2 g protein-day-1. Total ‘synthetic’ CO2 flux was determined during reproductive growth and a rate of ‘synthetic’ CO2 flux was determined during both reproductive and vegetative growth. Between 15 and 35 per cent of the CO2 fixed in the previous photoperiod was lost in ‘synthetic’ respiration of above-ground material in reproductive swards. Previous shading increased the proportion of ‘synthetic’ CO2 loss from above ground. The rate of ‘synthetic’ CO2 output during the first hours of darkness increased with amount of CO2 fixed in the previous photoperiod, although it was not proportional to it. There is some evidence that assimilate is ‘carried-over’ from one photoperiod to the next.