The Northeast China Transect (NECT) along a precipitation gradient was used to cal-culate the carbon balance of different vegetation types, land-use practices and temporal scales. NECT consists of mixed coniferous-broadleaved forest ecosystems, meadow steppe ecosystemsand typical steppe ecosystems. Analyses of the C budget were carried out with field measurement based on dark enclosed chamber techniques and alkali absorption methods, and the application of the CENTURY model. Results indicated that: (1) soil CO_2 flux had a strong diurnal and seasonal variation influenced by grassland type and land-use practices. However, the seasonal variation on soil CO_2 fluxes did not show obvious changes between non-grazing and grazing Leymus chinensis dominated grasslands. (2) Hourly soil CO_2 fluxes mainly depended on temperature, while dailyCO_2 fluxes were affected both by temperature and moisture. (3) NPP of the three typical ecosys-tems showed linear relationships with inter-annual precipitation, but total soil carbon of those eco-systems did not. NPP and total soil carbon values decreased westward with decreasing precipita-tion. (4) Model simulation of NPP and total soil carbon showed that mean annual precipitation was the major limiting factor for ecosystem productivity along NECT. (5) Mean annual carbon budget is the largest for the mixed coniferous- broadleaved forest ecosystem (503.2 gC m~(-2)a~(-1)), followed bythe meadow steppe ecosystem (227.1 gC m~(-2)a~(-1)), and the lowest being the typical steppe eco-system (175.8 gC m~(-2)a~(-1)). This study shows that concurrent field measurements of terrestrial ecosystems including the soil and plant systems with surface layer measurements along the wa-ter-driven IGBP-NECT are valuable in understanding the mechanisms driving the carbon cycle in different vegetation types under different land-use practices. Future transect research should be emphasized. The Northeast China Transect (NECT) along a precipitation gradient was used to cal-culate the carbon balance of different vegetation types, land-use practices and temporal scales. NECT consists of mixed coniferous-broadleaved forest ecosystems, meadow steppe ecosystems and typical steppe ecosystems. Analyzes of the C budget were carried out with field measurement based on dark enclosed chamber techniques and alkali absorption methods, and the application of the CENTURY model. Results indicated that: (1) soil CO_2 flux had a strong diurnal and seasonal variation influenced by grassland However, the seasonal variation on soil CO_2 fluxes did not show obvious changes between between non-grazing and grazing Leymus chinensis dominated grasslands. (2) Hourly soil CO_2 fluxes mainly depended on temperature, while dailyCO_2 fluxes were affected both by temperature and moisture. (3) NPP of the three typical ecosys-tems showed linear relationships with inter-annual precipitat ion, but total soil carbon of those eco-systems did not. NPP and total soil carbon values ​​decreased westward with decreasing precipita- tion. (4) Model simulation of NPP and total soil carbon showed that mean annual precipitation was the major limiting factor for ecosystem productivity along NECT. (5) Mean annual carbon budget is the largest for the mixed coniferous-broadleaved forest ecosystem (503.2 gC m -2 a -1), followed by the meadow steppe ecosystem (227.1 gC m ~ (-2) a -1), and the lowest being the typical steppe eco-system (175.8 gC m -2 -1). This study shows that concurrent field measurements of terrestrial ecosystems including the soil and plant systems with surface layer measurements along the wa-ter-driven IGBP-NECT are valuable in understanding the mechanisms driving the carbon cycle in different vegetation types under different land-use practices. Future transect research should be emphasized.