It is widely recognized that major causes of crime in the United States lie largely outside the direct reponsibility and resources of law enforcement and justice agencies, causes such as family abuse and neglect, school failure, poverty, unemployment, and ethnic disadvantages. Still, the effectiveness of particular criminal justice policy is determined by the degree to which it controls crime and reduces recidivism. Although criminal justice policy has in the past supported both the elements of rehabilitation and deterrence through punitive sanctions, the 1960's are viewed as a period that emphasized rehabilitation measures. Appraisals of this approach, however, have failed to show any significant positive effect from the rehabilitation programs implemented. While some have used these findings to disparage rehabilitation policy altogether, this is an extreme interpretation of the findings. Continued efforts to match treatment programs to offender types found to be most responsive to them are justified. Stiffer penalties and mandatory sentencing have recently become powerful influences in penal policy, although their effectiveness in controlling crime and reducing recidivism has not been empirically established. This strategy should be as carefully evaluated as was the rehabilitation model. An evaluation of the effectiveness of stiffer drug penalties on drug crime in New York is presented. Tabular data and 16 references are provided.