This article traces findings from four historical waves of research on sentencing disparities. Particular attention is given to changes in research methodologies and data sources, the social contexts within which research has been conducted, and the various forms that bias can manifest. It explores the change from findings of overt racial/ethnic disparities to more subtle, but still systematic, institutionalized biases. In so doing, the movement toward determinate sentencing is discussed, and the biases identified are partially explained by the need for the system to maintain legitimacy in the face of social change. (Author abstract)