Many studies regarding growth in diameter at breast height (D) in trees suffer from several problems, including heteroscedasticity, temporal autocorrelation and very low statistical adjustments. In growth ring studies, growth models are sometimes omitted, presenting only a mean curve or smoothings, while studies that use models often do not address the above mentioned problems. For these reasons, this paper proposes a new approach to the classical modeling of D = f(t), where t is age (years), using the logarithmic transformation of the relative growth rate ln(1/D)(dD/dt) = ln f(D, A),where A is the asymptote of D based on the differential growth rate model of von Bertalanffy. High statistically significant adjustments for Cedrela odorata (ME = 65%, model efficiency, ME, an analogous to R2 but for non-linear regressions), and Juglans neotropica (ME = 78%) were obtained and met all regression assumptions. These equations were integrated to obtain D = f(t) for both species, followed by self and independent validation. Based on these equations, different life history and silviculture traits were calculated for both species. This procedure does not appear to have been previously used in the study of tree growth.