The spatial distribution and levels of available plant nutrients (elements) in the soil can limit coffee yield and must be evaluated for effective crop management. Therefore, we analyzed spatial variability in yield and plant nutrients in the saturation extract of a clayey Oxisol cropped with fertigated coffee. The experiment was carried out on 14 hectares of coffee in Monte Carmelo, Minas Gerais, Brazil. Soil samples were collected (0 - 0.2m layer) at 61 regular grid points (spaced 50x50m) and used to determine plant nutrients in the saturation extract. Coffee yield was also determined at these points. Descriptive statistics were calculated for each variable and geostatistics were used to build a spatial variability model representing the physical attributes of the soil. Variographic analysis was performed using semivariograms. These showed that yield and soil chemistry varied throughout the study site. Thus, the maps generated from geostatistics can be useful tools for soil management in fertigated coffee crops.