Abstract: To test the effectiveness of GF-1 16 m remote sensing image in sampling survey of US soybean planting area, and solve the problem of agricultural remote sensing survey in China which mainly relies on foreign satellite image. At the 95% confidence level and error of less than 5% , the authors designed a simulated sampling test of 2013 with 6 sampling fractions, soybean planting area in sampling frame was used as zone index and divided sampling cells into 6 layers, total relative error and variable coefficient were used as evaluation indexes, the minimum sampling fraction was 1.8%. The test of 2014 adopted the same method of 2013 sampling test, but the sample was GF-1 16 m remote sensing image with the sampling fraction of 4.8%, sampling results had 5.1% error compared with the statistical data of USDA. The results showed that in the growth season of soybean in US, GF-1 16 m remote sensing image could ensure the covering power of soybean planting area sampling survey. The results could provide reference for the design of crop planting area sampling survey project.