This paper examines the estimation of the number of workers with disabilities in the Nusa Tenggara region using Sakernas 2024 data. The limited sample sizes in several districts lead to high sampling errors, necessitating a more reliable small-area statistical approach (Small Area Estimation). The unavailability of accurate small-area labor statistics for persons with disabilities hampers evidence-based regional development planning and inclusive policymaking. This study applies the Small Area Estimation (SAE) method using a Hierarchical Bayesian (HB) Poisson–Gamma model to handle count data with overdispersion—an approach that remains rarely applied in Indonesian labor statistics. The model is developed by integrating Sakernas data with auxiliary information from PODES and the Ministry of Education. Estimation is conducted through Bayesian inference using Markov Chain Monte Carlo (MCMC) simulation. The HB Poisson–Gamma model effectively reduces the Relative Standard Error (RSE) from an average of 44.6% in direct estimation to below 10% across 32 districts in Nusa Tenggara. These results demonstrate the model’s ability to improve data reliability and support inclusive employment policies aligned with regional development priorities.