Abstract: This dissertation uses modern theory and method of arable land resource management, intergrates knowledge of GIS technology, according to the data of the second general survey soil, replenish mass data for database through investigating on the spot, choose GIS software, has established Ding’an county’s arable land resources database with 1:50000 scale. On the base of these databases, we has choosed 12 essential evaluative factors, such as drainage capability, crop rotation system, parent material, soil erosion type, soil acidity, organic matter, total nitrogen, available phosphorus, available potassium, exchangeable magnesium, plow layer depth, obstacle layer type. It was obtained 2411 evaluation units by laying over soil map and fundamental farmland protection plot map or current landuse map. According as the theory of Fuzzy Mathematics, adopting Telphi Principle, Analytic Hierarchy Program and Addition Model to calculate the compositive index of arable land productivity, and applying the principle of equidistant grading to form the grade of arable land productivity. According this time survey of arable land Ding’an county mainly to paddy soil, we have looked out the area of was 66729.47hm2. Aggregate index was applied to grade the grade of Ding’an county arable land, grading result as follows: grade two the area was 2241.60hm2, grade three was 27261.27hm2, grade four was 26065.93hm2 and grade five was 11160.67hm2, sharing each percentage as 3.36%, 40.85%, 39.06% and 16.73%, respectively. There were not grade 1, grade 6, grade 7 and grade 8 arable land, there was greater area in grade 3 and grade 4 arable land. The arable land productivity was medium level in China in term of the productivity classify system of nation-wide arable land.