Accurate estimates of aboveground biomass (AGB) strongly depend on the suitability and precision of allometric models. Diospyros mespiliformis Hochst. ex A. DC. is a key component of most sub-Sahara agroforestry systems and, one of the most economically important trees in Africa. Despite its importance, very few scientific information exists regarding its biomass and carbon storage potential. In this study direct method was used to develop site-specific biomass models for D. mespiliformis tree components in Burkina Faso. Allometric models were developed for stem, branch and leaf biomass using data from 39 tree harvested in Sudanian savannas of Burkina Faso. Diameter at breast height (DBH), tree height, crown diameter (CD) and basal diameter (D20) were regressed on biomass component using non-linear models with DBH alone, and DBH in combination with height and/or CD as predictor variables. Carbon content was estimated for each tree component using the ash method. Allometric models differed between the experimental sites, except for branch biomass models. Site-specific models developed in this study exhibited good model fit and performance, with explained variance of 81–98%. Using models developed from other areas would have underestimated or overestimated biomass by between –72% and +98%. Carbon content in aboveground components of D. mespiliformis in Tiogo, Boulon and Tapoa-Boopo was 55.40% ± 1.50, 55.52% ± 1.06 and 55.63% ± 1.00, respectively, and did not vary significantly (P-value = 0.909). Site-specific models developed in this study are useful tool for estimating carbon stocks and can be used to accurately estimate tree components biomass in vegetation growing under similar conditions.