Two plant-soil systems. C-3 plant-'C-4 soil' (obtained from a grassland dominated by C-4 grasses) and C-4 plant-'C-3 soil' (obtained from a pasture dominated by C-3 grasses). were used in this study to monitor the dynamics of rhizosphere respiration (root-derived CO2-C) at different plant developmental stages. In C-3 plant-'C-4 soil' system. CO2-C derived from soybean roots increased significantly from vegetative stage (55.69 mg C d(-1) pot(-1)) to flowering stage (132.18 mg C d(-1) pot(-1)), and it declined thereafter (83.37-111.63 mg C d(-1) pot(-1)). However, no significant change of CO2-C derived from sunflower roots was observed at different plant developmental stages (67.05-77.66 mg C d(-1) pot(-1)). CO2-C derived from soybean (Glycine max (L) Merr) roots was significantly higher than that derived from sunflower (Helianthus annuus) roots except for that at vegetative stage. In C-4 plant 'C-3 soil' system. CO2-C derived from sorghum roots was significantly higher at flowering stage (169.51 mg C d(-1) pot(-1)) than at other stages (75.89-113.26 mg C d(-1) pot(-1)). C-2-C derived from amaranthus roots was the highest at vegetative stage (88.88 mg C d(-1) pot(-1)) and it declined significantly thereafter (23.42-53.47 mg C d(-1) pot(-1))-CO2-C derived from sorghum (Sorghum bicolor) roots was significantly higher than that front amaranthus (Amaranthus hypochondriacus) roots except for that at vegetative stage. In conclusion, rhizosphere respiration varied not only with plant species but also with plant phenology. With the soil volume used in our pots. the overall percentages of cumulative root-derived CO2-C to total soil respiration were 61.22, 61.14. 81.84 and 67.37% for soybean. sunflower. sorghum and amaranthus, respectively. Specific rhizosphere respiration was also discussed and was used as an index of root activity and vitality.