Diarrhea continues to pose a significant public health challenge in Makassar City, with incidence varying across sub-districts. Mapping diarrhea risk is essential for public health planning, as it helps identify high-risk areas and allocate resources efficiently. Accurate spatial risk assessment supports targeted interventions and informs evidence-based health policies. This study aimed to identify areas with high and low relative risks (RR) of diarrhea cases using Bayesian spatial Conditional Autoregressive (CAR) models, specifically the Besag–York–Mollié (BYM) and Leroux approaches. The analysis was based on case data from 15 sub-districts in Makassar City in 2023. Model performance was assessed using the Deviance Information Criterion (DIC) and the Watanabe–Akaike Information Criterion (WAIC). The CAR-Leroux model with an Inverse Gamma (IG) hyperprior (0.5; 0.0005) was identified as the best-fitting model, providing the most reliable estimation of relative risk. Kepulauan Sangkarrang exhibited the highest RR, indicating a markedly elevated risk of diarrhea relative to the city average, while Biringkanaya District showed the lowest RR, reflecting a substantially lower risk compared to the average.Keywords: Bayesian spasial models, CAR BYM, CAR Leroux, Diarrhea, Relative risk.