ABSTRACT: This study aimed to evaluate the spatial variability using the indicative kriging method in preparing probability maps of the occurrence of soil chemical properties and productivity. Data were collected in a ground area planted with soybeans in a no-till system, with a regular grid sampling of 60 x 60 m, totaling 124 points. At each point were deformed soil samples collected to determine the chemical attributes (pH, Potassium, Calcium, Magnesium and Aluminum exchangeable, available phosphorus, zinc, potential acidity, soil organic matter, cation exchange capacity and base saturation). To determine productivity were collected all the plants within 1 m² of the sample point. Data were categorized into binary codes on the amounts presented in the recommendation for the use of lime and fertilizer. Later it used the geostatistical analysis and attributes potassium, phosphorus, magnesium and base saturation that presented spatial dependence underwent kriging as interpolation technique. Occurrence probability maps were drawn of the evaluated attributes presented spatial dependence. The probabilistic maps showed that the central area evaluated was more likely to need correction to the levels of potassium attributes, phosphorus, magnesium and base saturation. From the analysis of the maps, you can detect nutrient deficient areas and making corrections and use of localized management techniques, minimizing operating costs and economic that culture.