<p>This study conducts a spatial econometric analysis of police-reported domestic violence against women in 2022 across the 52 communes of Chile’s Metropolitan Region. It examines the spatial distribution of violence and its association with population density and a synthetic education index (EDUC), constructed through principal component analysis. Results from Moran’s I tests confirm significant spatial autocorrelation in the variables, justifying the use of spatial models. Model comparison identifies the spatial lag model (SLM) as the most appropriate specification, revealing that domestic violence rates in one commune are influenced by neighboring areas. Estimates indicate that higher values of the EDUC index and greater population density are significantly associated with lower domestic violence rates. Moreover, spatial decomposition shows positive spillover effects: improvements in education and density in one commune can affect violence outcomes in adjacent communes. These findings highlight the importance of spatially coordinated strategies and the analytical value of using composite indicators like EDUC in understanding the geographic dynamics of gender-based violence.</p>