The number of data elements needed to reasonably array soil productivity indices is dependent upon the size of the geographical area. More data elements are needed as the geographical area increases in size. This is because: 1) more and different soils in an area require more soil data elements to properly array the soils, 2) regional scale temperature and rainfall variation affects yield so climate data are needed, and 3) the interaction of data elements for productivity indices between areas within large geographical areas are different. The soils information used to generate soil productivity indices is housed in the National Cooperative Soil Survey Soil Data Mart. The Soil Rating for Plant Growth (Version 4.0) computer program uses non-crisp evaluation criteria to determine the productivity indices. Earlier versions of Soil Rating for Plant Growth used crisp evaluation criteria. There are 21 non-crisp criteria submodels. Each has a value ranging from 0.01 to 1.0. The 21 submodel numbers are grouped into physical soil properties, chemical soil properties, landscape features, and climate. The numerical values for these four groups are multiplied together to get a product. This product is then multiplied by 100. The productivity indices for soils range from 5 to 100 (any soil with a productivity index of less than 5 is assigned a value of 5). The soil productivity indices can be grouped to generate thematic maps using digitized soil surveys. The thematic map in this paper shows three groups of soils: non-arable, very highly productive soils for commodity crops, and other productive soils.