Are there ways to intervene early in a child's life that might reduce, at a reasonable cost, the probability of his becoming a serious delinquent? The results of some small-scale social experiments have been published suggesting that certain kinds of preschool education and parent training might have desirable and lasting effects. In addition, there is growing evidence that some kinds of medical intervention and certain forms of school organization and ethos could help reduce the rate of misconduct. The authors provide a full-scale assessment of the evidence that might lead to the design of new research and action efforts.