To investigate the characteristics of spatial and temporal patterning of arson incidents in a large urban county.

Using multiyear geocoded crime data from the Los Angeles County Sheriff’s Office (2005 to 2012), arson incidents are analyzed for significant space–time interaction using the Monte Carlo simulation-based Knox method, originally developed for epidemiology research. Several additional contrasts are presented according to situational and routine activities factors that may be implicated in target selection and guardianship efficacy.

Evidence illustrates that arson generally exhibits enhanced likelihood of near repeat follow-up incidents in close spatial and temporal proximity to an initiating event. Also, spatiotemporal patterns for arson exhibit systematic differences when contrasting subtypes including residential versus non-residential, daytime versus nighttime, and weekday versus weekend arsons.

Comparatively little empirical research has been published in criminology journals concerning this "invisible" Uniform Crime Report index crime. This study provides a theoretically informed, aggregate-level examination of trends and patterns for arson in an urban environment. Results may aid in inductive theory building as well as efforts to investigate and prevent arson.