Abstract: The risk of severe illness or death from COVID-19 is associated with specific demographic characteristics or composition of the population within geographic areas, and the spatial relationship between these areas. The aim of this paper is to identify areas with a higher concentration of population vulnerable to COVID-19, relying on the concept of spatial dependence. Hence, we focus on the share of vulnerable populations using several salient proxy measures at municipality level data for Serbia. The degree of vulnerability at the municipality level was determined by hotspot analysis, specifically the Getis-Ord Gi* statistics. The results indicate heterogeneity in the spatial patterning and typologies of clusters across Serbia. This spatial heterogeneity reveals potentially differing degrees of risk across municipalities. The results can inform decision-makers in the fight against COVID-19 by helping to identify those areas with vulnerable populations that if exposed may stress the local health care system.