Abstarct: Soil Map is one of the most basic and comprehensive tools to study the soil types; therefore should be carefully considered in its preparation. Providing an updated and reliable soil map with the possibility of combining it with other data laxyers of economic-social models and environmental context, can greatly contribute to exact evaluation of soils and land conditions, predicting soil characteristics, water and other normal and abnormal factors. In recent years regarding to spatial predictive of soil types, digital soil mapping has noticeable increases in achieving significant environmental covariance and strength analysis. On the other hand, Google Earth provides capabilities of monitoring, management, processing and analysis of soil maps and other maps with great details. The ultimate goal of this system (for those whom the data and spatial information is greatly needed) is helping the spatial of decision making process. Also Google earth -as a powerful system and being always update- is used for showing changes in soils and publishing soil maps in making optimal data, information, time and wealth creation. In this study, with the help of decision tree method and Scorpan Model with Google Earth and GIS; we could provide and update the digital soil map of Tehran`s southwest. For this purpose, five factors of soil-Maker (climate, clothing and living organisms, topography, parent material and time) was used in the Scorpan Model (Scorpan, 2003) for inputting decision tree method; along with auxiliary spatial data such as land form, remote sensing, and digital evaluation model. Method of decision tree is a non-parametric approach with combination of data, knowledge baxse and computer inference; so that it is very useable in environmental sciences because of managing data and information. The obtained digital soil maps were validated by using existing cripple (customary) soil maps, and accuracy of the map was obtained in category 92% , sub-category 85% and 36% for large groups. The result of study indicated that updating soil maps with decision tree and using Google Earth and GIS with a low number of observations is possible and have reasonable accuracy and precision. Moreover, it can be easily combined with other information laxyers and published.