Geographic Information Systems (GISs) are frequently used to perform spatial analyses in many fields of study. Because most GIS tools in widespread use today are not capable of easily incorporating the element of time into an analysis, the majority of the results produced are either in the form of a single, static snapshot using data that have been aggregated for a period of time, or in the form of multiple snapshots representing different time periods for a comparative static analysis. For some analyses, these techniques may be appropriate. But in other analyses, such as a study of urban crime, these methods may result in the loss of important information about the conditions under which crime occurs. This research presents a new methodology that explicitly incorporates a high-resolution temporal dimension into a statistical analysis of urban violent crime. The methodology not only involves the disaggregation of daily crime incident data into appropriate space-time units, but also captures the concurrent changes in the population at risk. To estimate urban population levels at this temporal resolution, an innovative approach using a micro-simulation model of daily activity travel patterns is employed. As a demonstration of this methodology, incidents of aggravated assault in Dallas, Texas, during the year 2000 are analyzed and the results compared to a conventional aggregate analysis.