Abstract: Winter wheat is the staple summer grain in China. The accurate estimation of its acreage and production is important. Neither the traditional investigation method nor the full coverage remote sensing using TM images is applicable because of the expenditure. A sampling and scaling scheme, which is economically feasible and scientifically sound, was proposed here. Firstly, all of the 1411 winter wheat growing counties in China were stratified into 6 strata according to their statistical winter wheat acreage in the past decade. Then about 20 TM scenes were used for sampling. The winter wheat acreage change between two years was drawn with TM interpretation. The change for each sampled county was calculated. A scaling model is used to draw the acreage change of overall winter wheat for the whole country.