The chapter focuses on what the available data on crime victimization are capable of telling us about factors associated with the risk of being a victim of crime. The author advises that crime-victimization data are limited in explaining the causes of crime. There is no empirically certain means of knowing what factors caused the specific occurrence of the victimization event, because there is no way of reliably constructing factors that would have prevented the victimization from occurring. In other words, "we cannot evaluate the actual probability of the event in question occurring against the unactualized probability of it not occurring." The author notes that some compensation for this difficulty might be incorporated into research by building "controls" into the design and analysis of studies of observed data; for example, researchers might establish temporal order from which to infer victimization causation by using longitudinal "panel" designs (i.e., repeated measurement of the same people at different points in time) and/or to control for extraneous or confounding effects by using appropriate, multivariate statistical models. Unfortunately, longitudinal panel data are rarely available on victimization. The author advises that the only viable research strategy is to work with the observations of crime victimization that can be accessed, but to approach these events as a source of theory development that must be tested in appropriate ways. This requires careful attention to how the derived hypotheses are framed and tested, particularly regarding inference about what would have occurred if the variables in a given criminal incident were removed or altered. 2 tables, 1 figure, 19 notes, and 65 references