Height-diameter models are fundamental for the inventory of timber, biomass and carbon. Located in the central region of Mexico, Abies religiosa is a conifer with high timber value that lacks models to estimate its height as a function of diameter. The objective of this study was to determine a local height-diameter model for A. religiosa trees under a mixed-effects model approach in the region of Tlaxco, Tlaxcala, Mexico. The first step consisted in evaluating the quality of fit of three growth models and one power model. We used height-diameter data of 1,539 trees measured in 197 circular plots of 1,000 m². The selected model was the Weibull, which was adjusted in a second step under the mixed-effects modelling approach. Our results showed that the most accurate predictive capacity was obtained when using the sub-basin covariate as a clustering factor and the random effect in the parameter corresponding to the asymptote. Under this approach, the goodness of fit of the model was superior to the traditional one adjusted by nonlinear least square. Comparatively the gains in fit statistics R², RSME and AIC were of 12.19 %, 47.99 % and 8.46 %, respectively. The proposed model is robust, reliable, and biologically coherent for its operational application in the forest management of A. religiosa in the study region.