Abstract: Agro-ecosystem is one of the terrestrial ecosystems under intensive control and human disturbance. The estimation of carbon (C) source or sink in agro-ecosystems is a focal research in global C-cycle studies. As carbon dioxide (CO₂) is generally the main greenhouse gas with significant effect on climate change, there has been a growing interest in analyzing and understanding C-flux from agro-ecosystems as affected by regional environmental conditions. To determine diurnal and seasonal variations in net ecosystem exchange (NEE) and to explore the effects of environmental factors, CO₂ flux was contAgro-ecosystem is one of the terrestrial ecosystems under intensive control and human disturbance. The estimation of carbon (C) source or sink in agro-ecosystems is a focal research in global C-cycle studies. As carbon dioxide (CO₂) is generally the main greenhouse gas with significant effect on climate change, there has been a growing interest in analyzing and understanding C-flux from agro-ecosystems as affected by regional environmental conditions. To determine diurnal and seasonal variations in net ecosystem exchange (NEE) and to explore the effects of environmental factors, CO₂ flux was continuously measured in 20112012 using the eddy covariance technique in rice-wheat rotation system in the Yangtze River Delta. During the study, eddy covariance measurement together with measurements of various soil and meteorological conditions were taken for two full growing seasons per year. To derive complete time series of NEE, flux partitioning and gap-filling methods were devised. The results showed significant trends of monthly average diurnal and seasonal variations in NEE in rice-wheat rotation system with a large C-sequestration capacity. Monthly average diurnal variations in NEE for different months depicted a U-shaped curve with varying peak values, the maximum peaks appearing at about 12:00 at noon. Seasonal variation in NEE tracked a W-shaped curve in the year for the two crops (winter wheat and summer rice). Maximum daily net CO₂ uptake reached 1.12 mg·m²·s^-1 in April for wheat and 1.45 mg·m²·s^-1 in August for rice. At the same time, daily cumulative C-uptake of wheat and rice reached the maximum values of 12.88 g(C)·m²·d^-1 and 10.63 g(C)·m²·d^-1, respectively. The rice-wheat rotation system in the Yangtze River Delta was a strong C-sink, with annual carbon fixation of 769.61 g(C)·m²·a^-1 for the period 2011–2012. On the whole, the characteristics of NEE were closely related with crop growth and meteorological conditions. The environmental factors influencing NEE at daytime were different from those at nighttime for both winter wheat and summer rice. The main environmental factor impacting NEE was photosynthetically active radiation (PAR) during the daytime and the relationship between PAR and daytime NEE during the two crop growing seasons was well represented by Michaelis-Menten Equation (R² = 0.370.83). NEE increased with rising temperature and PAR, and decreased when PAR exceeded 1 800 μmol·m²·s^-1. Temperature was identified as the main environmental factor influencing NEE at nighttime. There was significant exponential correlation between nighttime NEE and temperature at different levels (air temperature, soil temperatures at 10 cm, 20 cm, and 40 cm depths) in rice-wheat rotation system in the Yangtze River Delta. However, the most related temperature level for nighttime NEE was driven by climatic conditions and crop growth. Correlation analysis of nighttime NEE and temperature suggested that 10 cm depth soil temperature was the most related for winter wheat and air temperature most related for summer rice. To further explore the relationship between temperature and nighttime NEE, there was need for monitoring nighttime fluxes with the combined use of eddy covariance and chamber-based method.