In 1990, the US Supreme Court in the case of Michigan Department of State Police v. Sitz declared sobriety checkpoints constitutional, in part because they were effective in reducing drunken driving. This article reviews the scientific literature on checkpoint effectiveness in order to evaluate the claim of deterrent accomplishments made in the brief supporting the police. It concludes that studies available at the time of the initial trial were insufficient to support the claim, but that more recent studies lay a better foundation for believing that checkpoints can deter drunken driving. The Court's definition of effectiveness, however, turns out to have been based on checkpoints' yield of arrests rather than on considerations of deterrence. The Court did not rely on the effect of checkpoints in increasing the perception of a deterrent threat or their accomplishments in actually reducing highway casualties in support of its conclusion. References, cases cited