Abstract: Abstract: It has a great significance on understanding the growth process of corn and guiding the agricultural irrigation to reveal the patters of corn water consumption, carbon assimilation and ecosystem water use efficiency. Therefore, this research collected flux and meteorological observations from 1997 to 2005 from three flux observations sites of the AmeriFlux network; and these three sites, focusing on farmland ecosystem, were located at corn fields in the northeast of the United States of America. At these three sites, latent heat, sensible heat and carbon dioxide fluxes were measured by eddy correlation system, and at the same time, other climatic variables, such as air temperature, vapor pressure deficit, and photosynthetic active radiation were also recorded by automatic meteorological stations. Based on the observations, this research estimated the water consumption (evapotranspiration, ET, including soil evaporation and corn transpiration) according to the latent heat flux, gross primary productivity (GPP) and net ecosystem productivity (NEE), both of which were estimated by according to the carbon dioxide flux; and then estimated the water use efficiency at different time scales, namely daily, weekly and annual, and furthermore tried to explore the relationship of water use efficiency with vapor pressure deficit (VPD) and photosynthetically active radiation (PAR) by using a correlation analysis. The results showed that the growth duration of corn is 156-180 days, ET ranges from 440 to 520 mm, GPP ranges from 1320 to 1640 g/m2, and NEE ranges from -560 to -620 g/m2; the peaks of weekly average ET, NEE and GPP are 3.5-5.3 mm/d, -10.0 - -12.5 g/(m2 d) and 17.5-28.0 g/(m2 d) for the three sites. It could be found that air temperature had a notable impact on corn growth duration; and specially, air temperature at the US-Ro1 site was 2℃ warmer than that at the other sites, while the corn growth duration at the US-Ro1 site was about 14 d longer than that at the other sites. Regarding the water use efficiency, WN (defined as NEE/ET) ranges from 1.2 to 1.4 g/kg, WG (defined as GPP/ET) ranges from 3.0 to 3.5 g/kg at an annual scale. Both WG and WN had a small variation at an inter-annual scale, with a variation coefficient of 0.2; however, at an intra-annual scale, they had a large variation, with a maximum of 4.5-5.7 g/kg in WG and 2.4-3.2 g/kg in WN, and the minimums close to zero, which was synchronous with GPP variation. More remarkably, WG had a negative correlation with vapor pressure deficit (VPD) at a canopy scale, which could be explained in theory according to the Farquhar's photosynthesis model. In this research, a more accurate relationship between ET and VPD was revealed, i.e. there was a constant k that made ET and GPP*VPDk approximately keep a direct proportion, with k value ranging from 0.42 to 0.63 for the three sites, which was close to the value of 0.5 proposed in the previous research. In addition, water use efficiency had no significant correlation with photosynthetic active radiation.