The authors' goal in this article is to contribute conceptually and empirically to assessments of the racial invariance hypothesis, which posits that structural disadvantage predicts violent crime in the same way for all racial and ethnic groups. Conceptually, the authors elucidate the scope of the racial invariance hypothesis and clarify the criteria used for evaluating it. Empirically, the authors use 1999-2001 averaged arrest data from California and New York to extend analyses of the invariance hypothesis within the context of the scope and definitional issues raised in our conceptual framing - most notably by including Hispanic comparisons with Blacks and Whites, by examining the invariance assumption for homicide as well as the violent crime index, by using discrete as well as composite disadvantage measures, and by using census place localities as the study unit. The mixed findings the authors report from their comparisons (across Whites, Black, and Hispanics; offense types; and type of disadvantage) suggest caution and uncertainty about the notion that structural sources of violence affect racial/ethnic groups in uniform ways. The authors conclude that the hypothesis should be regarded as provisional, and its scope remains to be established as to whether it applies only under narrow conditions or is a principle of general applicability. Tables, references, and appendix (Published Abstract)