Nowadays, the culture of the sugarcane plays an important role regarding the Brazilian reality, especially in the aspect related to the alternative energy sources. In 2009, the municipality of Suzanapolis (SP), in the Brazilian Cerrado, an experiment was conducted with the culture of the sugarcane in a Red eutrophic, with the aim of selecting, using Pearson correlation coefficients, modeling, simple, linear and multiple regressions and spatial correlation, and also the best technological and productive components, to explain the variability of the productivity of the sugarcane. The geostatistical grid was installed in order to collect the data, with 120 sampling points, in an area of 14.53 ha. For the simple linear regressions, the plants population is the component of production that presents the best quadratic correlation with the productivity of the sugarcane, given by: PRO = -0.553**xPOP(2)+16.14*xPOP-15.77. However, for multiple linear regressions, the equation PRO = -21.11+4.92xPOP**+0.76xPUR** is the one that best presents in order to estimate that productivity. Spatially, the best correlation with yield of the sugarcane is also determined by the component of the production population of plants.