Research has shown that mapping techniques are useful in forecasting future crime events. However, the majority of prospective mapping techniques has focused on the event-dependent influence of instigator incidents on subsequent incidents and does not explicitly incorporate the risk heterogeneity of the setting. The study here discussed is a modest attempt to address this issue by using a two-step process: first, using risk terrain modeling, we operationalized the “environmental backcloth,” (the risk heterogeneity of an area) to forecast locations of residential burglaries in the urban city of Newark, New Jersey. Second, using the near repeat calculator, we assessed the variability of underlying risk between different types of residential burglaries. A discussion of the findings and the joint utility of these approaches is provided.