This study investigates the concept of criminality of place, specifically considering residential theft (i.e., residential burglary and residential larceny). By utilizing official police records, this study examined the spatial distribution of theft across residential locations in a large, Midwestern city. Land parcel level analyses were conducted using geographic information systems (GIS), tabular analyses, logistic regression, and a spatial-scan statistic. The results showed that not only were thefts not uniformly distributed among residential locations, but that a few locations (less than five percent) were responsible for most of the thefts reported to the police. Further, the spatial-scan statistic analysis indicated that that risk of residential theft is significantly clustered at relatively few places and the non-random distribution is likely due to some underlying risk factors. Finally, this study then examined the characteristics of these locations to determine whether elements of social disorganization, routine activities or an integrated theory perspective resulted in the prediction of these high theft incident locations. The results showed that a combined model had more predictive ability than either theory alone.