Throughout the years, crime analysts continue to regard the incorporation of spatial analytics in crime analyses. This interest, alongside various technological advancements with geographic information systems (GIS), in crime visualization plays a vital role in understanding crime dynamics with the intention of aiding in strategic planning. Utilizing Empirical Bayes smoothed crime rates (per 100,000 population) in 2010, this paper aims to identify the presence of spatial autocorrelation within index crime rates observed in the National Capital Region (NCR) using Moran’s I. Spatial units were taken to be the 16 cities and one municipality in NCR. Results revealed that among the seven index crimes, only two, murder and physical injury, were able to exhibit significant spatial autocorrelation. Accounting for the presence of spatial autocorrelation, a regression model which takes spatial dependence into account was used. In particular, spatial lag models were fitted to study the relationships of murder and physical injury with certain demographic variables across cities. It was observed that population density and percentage of male individuals aged 15-24 are found to be positively correlated to the incidence of both of these crimes. On the other hand, factors such as the percentage of married adult population, percentage of foreign immigrants, percentage of the population with a high school diploma, and the crime solution efficiency correlate negatively to the incidence of these crimes. These observed factors could serve as indicators of crime incidences and, thus, monitoring them could aid in crime prevention and security management.