To control the growth rate of diarrheal cases, one of the steps that can be taken is to analyze factors related to diarrhea and their potential to contribute to an increase in diarrhea cases from a regional perspective. Understanding the influence of regions or geographic factors in modeling is known as spatial modeling. It has been identified that there are two common variants of spatial models often used, which include the Spatial Autoregressive Model (SAR) and the Spatial Error Model (SEM). The objective of this research is to determine the best model for the incidence of diarrhea in the Bali region by applying spatial analysis methods such as SAR and SEM. The research results indicate that the data used in the analysis contain significant spatial elements, highlighting the importance of integrating spatial aspects into modeling. In this context, the SEM method has proven to be the most suitable model when compared to Ordinary Least Squares (OLS) and SAR. The R-squared value for the SEM model is 96.01%, and it achieves the lowest AIC value of 267.558. Therefore, these findings demonstrate that the SEM model is more accurate in explaining the relationships among the variables involved in this study.