The use of allometric equations in the estimation of the commercial volume of wood allows to plan the management Forestry In that sense, the objective in this work was to select the best allometric models to estimate the volume of wood in G. crinita trees, for this, trees between 13.09 m high and 10.58 cm DAP were used on average. For the determination of the best model, the highest adjusted coefficient of determination R2 (> R2Aj), lower coefficient of variability (< CV%) and lower Furnival index (IF) were considered. Of the 17 existing allometric models for forest species, two were suitable, the model: ln (V) = a + b ln (D) (R2 = 0.93, CV = 11.93%, IF = 0.003444) and the model that uses the combined variable Log: ln (V) = a + b ln (D2*H) (R2 = 0.91, CV = 11.93%, IF = 0.003444), likewise, a third model proposed in this study: ln (V) = a + b ln (D2) + c ln (D * H) - d ln (D2*H) (R2 = 0.98, CV = 11.93%, IF = 0.003444), also presented the best fit to estimate the commercial volume of G. crinita at 31 months of age. However, the use of the Husch model is recommended, as it is easy to apply, since it only uses the diameter at chest height (DBH) as an independent variable.