Agroforestry offers a huge potential for carbon sequestration, contributing to climate change mitigation and carbon trading. This study focuses on Bauhinia variegata and Celtis australis, two important agroforestry tree species in the Western Himalayas aiming to develop allometric models and biomass prediction ratios using empirical data collected through selective sampling and minimally destructive methods. Biomass components were categorized and weighed, and allometric equations were developed using diameter at breast height and height as independent variables. Model stability was validated using cross-validation techniques, and their predictive accuracy was assessed. Models, particularly based on diameter, has significant predictive ability for predicting the biomass components for both the species. B. variegata demonstrated a higher capacity for CO