Effects of elevated C 0 2 (525 and 700 #L L-l), and a control (350 #L L -1 CO2), on biochemical properties of a Mollic Psammaquent soil in a well-established pasture of C3 and C4 grasses and clover were investigated with continuously moist turves in growth chambers over four consecutive seasonal temperature regimes from spring to winter inclusive. After a further 'spring' period, half of the turves under 350 and 700 #L L -1 were subjected to 'summer' drying and were then re-wetted before a further 'autumn' period; the remaining turves were kept continuously moist throughout these additional three consecutive 'seasons'. The continuously moist turves were then pulse-labelled with 14C-CO2 to follow C pathways in the plant/soil system during 35 days.Growth rates of herbage during the first four 'seasons' averaged 4.6 g m -2 day -1 under 700 #L L -1 CO2 and were about 10% higher than under the other two treatments. Below-ground net productivity at the end of these 'seasons' averaged 465, 800 and 824 g m -2 in the control, 525 and 700 #L L -l treatments, respectively.In continuously moist soil, elevated CO2 had no overall effects on total, extractable or microbial C and N, or invertase activity, but resulted in increased CO2-C production from soil, and from added herbage during the initial stages of decomposition over 21 days; rates of root decomposition were unaffected. CO2 produced h -1 mg -l microbial C was about 10% higher in the 700 #L L -1 CO2 treatment than in the other two treatments. Elevated CO2 had no clearly defined effects on N availability, or on the net N mineralization of added herbage.In the labelling experiment, relatively more 14C in the plant/soil system occurred below ground under elevated CO2, with enhanced turnover of !4C also being suggested.Drying increased levels of extractable C and organic-N, but decreased mineral-N concentrations; it had no effect on microbial C, but resulted in lowered microbial N in the control only. In soil that had been previously 'summer'-dried, CO2 production was again higher, but net N mineralization was lower, under elevated CO2 than in the control after 'autumn' pasture growth.Over the trial period of 422 days, elevated CO2 generally appears to have had a greater effect on soil C turnover than on soil C pools in this pasture ecosystem.