This model is derived from the literature on deterrence, as well as assumptions made about how RBT would affect drunk driving behavior; the original model has been modified in light of the results of empirical research conducted in New South Wales and other States, which has shown that RBT has had a causal impact on fatal automobile crashes. The main strength of the model is that it is grounded in a specific ongoing legal intervention that may be implemented in various forms, but which always targets the single offense of drunk driving. Because the model is offense-specific, it is possible to develop detailed concepts and measures that tend to be ignored in more general theoretical formulations of deterrence. A comparison of this model with similar models for other offenses can make a significant contribution to the development of a general theory of deterrence. 2 figures, 4 notes, and 49 references