According to most reports, the United States has experienced a steady decline in intimate partner homicide since 1976. At first blush these declining rates would seem to reflect the effectiveness of criminal justice strategies to reduce intimate partner violence. However, careful analysis of the data indicates that the plunge in intimate partner homicide is due mainly to the reduction in male victimization rates, which are also marked by racial differences. The author disaggregated rates of victimization by gender, ethnicity, and martial status in order to offer a detailed measure of victim safety and a precise method for developing theoretical explanations for the drop in intimate partner homicide. A synthesis of recent empirical research regarding policy-relevant factors that have been hypothesized to increase victim safety is presented. The analysis focuses on tests of exposure reduction and backlash hypotheses in regards to the reduction in intimate partner homicide. The overall analysis indicates that while in the aggregate, intimate partner homicide has declined, the benefits of increased victim safety are not equally distributed across all classes of victims. The main policy implication is that different classes of victims require different types of interventions and that these interventions should reach all victims. Future research should examine the aggregate outcomes of criminal justice policies designed to reduce intimate partner violence. Note, references