Models of racial attitudes traditionally have assumed that individual differences in the strength of underlying, 'implicit' associations between racial categories and stereotypical traits are the primary determinant of the expression of race bias. Thus, individual differences in performance on laboratory tasks designed to assess implicit race bias tend to be interpreted in terms of association strength. Here, we argue that such associations tell only part of the story, and probably the least interesting part. We posit that response conflict and its regulation are critical to understanding the need for control, and that affect-related processes help to determine the extent to which control resources will be implemented to overcome biased associations. We present data from a number of recent behavioral and psychophysiological studies in support of this idea, as well as conceptual accounts that point toward a model of race bias regulation that depends upon processes identified as important for regulation of thought, affect and action more generally.