Agricultural water productivity is an important indicator to assess the effect of irrigation water management and water conservation.In this paper

a method is presented to calculate water productivity based on evapotranspiration（ET） and dry matter（DM） estimated by remote sensing data.Water productivity data are calculated and analyzed in the plain of Hai Basin from 2003 to 2008.The results show that water productivity data estimated by remote sensing can reflect spatial differences in the amount of water output per cubic meter.The average water productivity for the plain of the Hai Basin is 0.99 kg/m3

the range is 0.02—2.13 kg/m3

and the overall level is low with a big difference between north and south.The relationship between yield and water productivity shows a high linear correlation（R=0.97）

and increasing yield production remains an important direction for improving the water productivity level in the basin.The curves between water productivity change with ET and yield changes with ET are similar

keeping stable after a linear increasing trend.When the water consumption of the wheat-and-maize rotation region is between 335—575 mm

water is the main limiting factor for yield production.When water consumption is higher than 575 mm

the harvest index is another important limiting factor for yield production except available water.