Most recent efforts to reduce drunk driving have been based on an explicit or implicit theory of simple deterrence, which holds that if the cost of an undesirable behavior can be increased, the behavior will decrease. Some strategies for increasing deterrence are to increase the probability of apprehension and conviction; increase processing costs (e.g., detention until sober); and expand the use of jail, fines, license suspensions and revocations, community service, public condemnation, vehicle impoundment, forfeiture, house arrest, and electronic surveillance. Obstacles to deterrence effectiveness in the case of drunk driving are that the chronic, dangerous offenders are not likely to calculate the costs of offending. If the high risk of personal injury or death as well as possible injury to their families and friends is not sufficient to deter drunk drivers, increased criminal sanctions are not likely to be effective. The most important role of deterrence strategies may be symbolic and long-term, contributing over time to new societal norms regarding the seriousness of drunk driving. This will serve to increase internal inhibitions against such behavior.