Studies show that heterogeneity among plants interferes with stand productivity, however, this information has not been incorporated in models to assist forest managers. This paper presents a new format for growth modeling, that includes the effect of heterogeneity on final stand productivity and was developed for unthinned and thinned Pinus taeda stands. The data used came from a 3 x 3 factorial experiment with the factors initial density (2500, 1250, and 625 trees.ha− 1) and thinning (without thinning, moderate thinning, and heavy thinning). The diameter distribution represented by the percentile method were used to represent stand homogeneity. The diameters (located in the 10th and 63rd percentiles) were inserted into the basal area growth model, reducing the mean absolute error (MAE) and the square root of the mean error (RMSE) on average from 4.8433 m².ha− 1 to 2.7702 m².ha− 1 relative to the predicted estimates, and from 4.3139 m².ha− 1 to 2.6984 m².ha− 1 for the projected estimates. The validation of the equation with the homogeneity proxy variable was performed by the Bootstrap method. A simultaneous equation, compatible in prediction and projection, with the inclusion of homogeneity, is recommended for estimating the growth in the basal area of Pinus taeda stands in southern Brazil.