A systematic sampling directed with geo-referenced points spaced every 100 m or so was used to evaluate the spatial variability of sand (%Sa), silt (%Si), clay (%C), organic matter (%OM), and pH in soils of the experimental unit The Glorieta for agricultural purposes. Classical statistics was applied to build semivariograms geo-statistical which were used by "Kriging" interpolation to characterize the variability of the sector. The spatial variation of the properties evaluated was from medium to low, with a spatial dependence between 347.3 and 469.7 m and a variance random between 0 and 77.83%. There was a strong spatial dependence in %L, pH and %OM, moderate to weak for %Sa and %C. The soil particle size showed greater variation from east to west, while pH and %OM varied more from north-south. These results provide a basis for experimental site location and selection of appropriate crops for the farm.