In the management of spatial variability of the fields, the management zone approach (MZs) divides the area into sub-regions of minimal soil and plant variability, which have maximum homogeneity of topography and soil conditions, so that these MZs must lead to the same potential yield. Farmers have experience of which areas of a field have high and low yields, and the use of this knowledge base can allow the identification of MZs in a field based on production history. The objective of this study was to evaluate the use of the farmer's experience in the delineation of MZs. Using data of elevation, soil penetration resistance, sand, silt, clay and soybean yield, the spatial correlation matrix was created to select the layers that influenced the yield. Then, the selected data were interpolated and the MZs were delineated using Fuzzy C-Means clustering method with SDUM (Software for Delineation of Management Zones). The farmer's experience layer was obtained using a mobile application developed for this. The MZs were delineated considering three cases: a) without the use of variable farmer’s experience; b) with the variable farmer’s experience and stable soil properties selected in the variable selection stage; and c) only with the variable farmer’s experience, considering two, three and four sub-regions. The study showed that the use of the farmer’s experience to set MZs can be an efficient and simple tool and reduce costs in the MZs setting process when compared to the traditional method of using stable soil variables and the relief. It should be noted that the good results obtained using the farmer's experience variable may have been positively influenced by the farmer's knowledge of this area for a long time.