In a flat lot of 4 ha, with dry warm climate and Mollisols were determined the bulk density, pH and content of sand, silt, clay, organic matter, Ca, Mg, K and P, in the first 150mm of soil, doing a sampling in a semi-regular grid with cell size of 25 x 25m. A factorial analysis was reduced the number of variables to two (factors). The contents of clay, Ca and Mg shaped the first factor, and the second the pH and the contents of organic matter and phosphorus. To each factor and their variables was made an analysis of semivariance and the respective map of its spatial distribution using interpolation with block kriging. In the map of factor 1 homogeneous areas were better defined, and there was a greater spatial correlation between this map and those of their variables. The factor 2 variables presented a spatial structure in patches that not corresponded with the factor structure, making its zoning produced not adequate to establish a possible programme of fertility management with techniques of precision agriculture. The zoning which gives the factor 1 should be used for this purpose.