ABSTRACT This study aims to generate the diagnosis and spatial variability maps of soil fertility attributes, as well as teak (Tectona grandis L.f.) production through geostatistics. The study was conducted on a commercial farm in the municipality of Capitão Poço, state of Pará, Brazil. Soil and yield data were collected from 155 sampled georeferenced points and 143 were used in the study (after outlier removal). The collected soil samples were submitted to laboratory analysis to obtain the values of the following variables: pH, organic matter (OM), calcium (Ca), magnesium (Mg) and potassium (K), available phosphorus (P), aluminum (Al), Base Sum (BS), cation exchange capacity at pH 7.0 (T), effective cation exchange capacity (t), base saturation (V%) and aluminum saturation (m %). Thereafter a diagnosis, linear correlation univariate and geostatistical analysis were applied in the resulting data. The univariate statistics showed that normal distribution is not required when evaluating spatial variability of chemical and production variables. Soil fertility diagnosis showed K, OM and P as the most limiting parameters in the commercial teak plot and the importance to fertilize forest areas. Positive correlation was found between P, OM, K with volume per tree. All soil fertility and teak volume variables showed spatial dependence, which enabled the production of spatial variability maps. The variability maps showed to be used complementary with univariate statistics to enable more precise interventions in a teak production area, as it showed the shortage of K in the area and the relation of P and volume per tree.