This paper contends that we know now to deter drunk driving, and nave done so in some States, to a degree. The systems that work are reviewed, and those that are counterproductive are critiqued. Actual case studies are used to support the theory that the legal and liquor industries are the major roadblocks to progress. Basic components of an effective deterrent system, and the strategies needed to put them in place, are outlined in some detail. Current data on road deaths, and the changing profiles of drunk drivers, their victims and their punishments are reviewed. The U.S. is going down the right road with a deterrence model but lacks a few key parts in the transportation system.