For the past century, most correctional treatment has emphasized rehabilitation rather than punishment. Programs have been based upon the medical model of crime as a symptom of individual pathology which may be cured by correctional experts. One criticism of rehabilitation is that it leads to the delegation of wide discretionary powers to criminal justice officials in the form of indeterminate sentences and parole, powers which are then used to control inmates. A review of the evaluative research of rehabilitative programs suggests that treatment programs are seldom successful. However, rehabilitative efforts often seem to fail because most forms of help are nonvoluntary, and the prison cannot remove those inmates who fail to meet minimal performance requirements. Programs are often judged on the basis of recidivism, and little research involves offenders who are released and then remain out of prison. Furthermore, many studies of rehabilitative programs are criticized for their poor empirical methodology, but controlled methods are extremely difficult to maintain in prison studies and therefore conclusions must be viewed with extreme caution. Preoccupation with practical problems and 'quickie' solutions to crime control thus should not cloud the importance of continuing basic research in the causes of crime. Seventy-six bibliographic footnotes are included.