Parameterizing photosynthetic capacity (Amax) associated with the lifetime carbon gain of tree species could aid in mitigating uncertainties within global terrestrial biosphere model (TBM) predictions. Due to the complexity of field conditions and limitations of apparatus, it is technically difficult and often labor- and material-intensive to measure the photosynthetic capacity of multiple tree species on a large scale of forest dynamics plots, especially canopy species. Given the extensive accessibility of global forest dynamics monitoring data, comprehensive research on allometric relationships, and multiple modeling studies exploring the correlation between tree growth and photosynthetic capacity parameters, utilizing tree dynamics census data to estimate the photosynthetic capacity of tree species in relation to their lifetime carbon gain is a promising approach. In this study, we present a novel framework that utilizes the Hierarchical Approximate Bayesian Computation method to estimate species-specific Amax, integrating it with a tree growth model based on allometric relationships. We illustrate our approach by estimating the photosynthetic capacities of 153 tree species in the 50-ha Forest Dynamics Plot on Barro Colorado Island, Panama. We assessed our approach by employing existing field-measured data on the photosynthetic capacity of 25 tree species. Our method yielded Amax estimation ranging from 3.35 to 31.30 μ mol m^-2 s^-1, with a Mean Absolute Percentage Error (MAPE) of 29.33%, indicating that our approach provides reasonable estimations across an empirical range of photosynthetic capacities. We also observed negative correlations between the photosynthetic capacity and allometric relationships of tree crown area and tree height. This study demonstrated the efficiency of using ABC to estimate Amax from tree census data and provided a promising pathway to estimate parameters in mechanistic models from empirical data. Our approach could potentially provide a robust foundation for parameterizing in vegetation models, which may help reduce uncertainties in TBM predictions.