Abstract The objective of this work was to evaluate the Huber (H), Smalian (S) and Newton (N) formulas in the scaling of Corimbya citriodora trees. Initially, based on the analysis of three different sections of the trunk, below and above 1.3 m, five methods of absolute cubage were developed using the H, S and N formulas. In the evaluations carried out, statistical criteria were adopted, such as the mean error. The design used was randomized blocks followed by the Dunnett test. From the adjustment of volumetric equations, the following statistics were adopted: adjusted coefficient of determination, residual standard error, mean squared error, mean percentage error, mean absolute error and sum of squares of relative errors. The results revealed that the H and S formulas resulted in an error of 3.43 and 4.29 times greater than the error made with the N formula, respectively. In addition, an excellent and expressive level of accuracy was detected for the N formula in relation to the H and S formulas, the latter being the one that presented the worst results. In estimating volumetric equations, there were no expressive differences between the H, S and N formulas, however, the faster implementation of the H and N formulas, combined with the level of accuracy, revealed how inadequate the use of the S formula is in sizing trees.