Abstract: Catalogue sampling has been used in conventional estimation of crop areas. But it is restricted by precision and efficiency because of the absence of realistic and effective transcendental knowledge. By combining random sampling, systematic sampling and stratified sampling, three new spatial sampling schemes were designed for crop area estimation based on the technologies of Geographic Information System, Remote Sensing and Global Positioning System. The differences of these three spatial sampling methods were synthetically analyzed on precision, stability and the smallest sample size. The experimental results show that spatial stratified sampling needs fewer samples and has higher estimation precision and stability. So this method can be popularized in the estimation of large scale crop areas.