Resumo. This work aims at evaluating a methodology in order to estimate the volume of wood in the forest inventory considering the stem taper quality of trees. A stand of 4.31 ha was used with an area planted with Eucalyptus sp. with sixteen years-old, originated from seminific propagation. The initial spacing used in the planting was 3 m x 3 m, being made a selective thinning at 8.6 years-old. Eight circular samples of 855 m² were located. In those samples, the diameter in the height of 1.30 m (DBH) and the total height (HT) of the trees were measured, being the tree stems DBH larger than 28 cm classified in agreement with a priori defined quality classes. The tree stems of the eight samples were submitted to the optimization process. After the crop of all trees, the stems were marked by a trained team and were sawed to subsequent comparison to commercial volume obtained by the optimization method. The volume and the number of logs for quality classes and assortment, optimized and harvested, differed significantly from the qui-square test to 5% of probability. It was concluded that it is needed standardization of the exploration and inventory team training, and the presented inventory methodology should be improved before implementation.