One of the issues that has influenced correctional policy since the 1970s has been the evaluation of the outcome of correctional treatment programs. While the public supports the concept of rehabilitation and correctional officials value them as effective means of changing inmates, there should at least exist an expectation that those programs will reduce recidivism. The ‘nothing works doctrine’, derived from Martinson's article, contributed in the disappearance of expectation and faith for correctional treatment programs. The reasons why Martinson's article of 'nothing works' have been so influential to academic as well as societal and political areas are explained in various respects: social and political climates of the 1960s-1970s characterized by anti-war moods, racial problems, poverty issues, and women's liberation, the changes of middle-class & liberal criminologists' attitudes towards crime and crime control, the politicization of crime problems, the domination of sociological approaches in criminology, and the lack of knowledge and skills in crime predictions by academics. In the US, rehabilitation as a correctional ideology had been retreating since early 1970s and almost disappeared in the mid 1980s as a goal of punishment, indicating that the retribution paradigm had returned. It is persuasive that such shifts resulted from political and societal reactions of conservatives and liberals toward crime, and are shown in correctional areas. In this context, some scholars has said that the society must move beyond the naivete and exuberance that marked the advocacy of rehabilitation in the 1950s and early 1960s and beyond the cynicism and pessimism that has reigned for much of the last three decades.(Palmer, 1992) This notion has received widespread support. New evaluation studies, including meta-analyses done after Martinson's article in 1974, indicate that correctional treatment programs could be effective in reducing criminal recidivism. And they have demonstrated that juvenile correctional intervention is more effective than intervention programs designed for adults. It has been known that behavioral/cognitive treatment, on average, produces larger effects than other treatment. Intensive, in-prison drug treatment is effective, especially when combined with community aftercare. Education, vocational training, and prison labor programs have modest effects on reducing criminal recidivism and increase positive behavior in prison. Evidence on sex offender treatment intervention program is less positive, probably because the target population is heterogeneous and treatment needs to be tailored to specific offender deficits.(Gaes et al., 1999) The results of meta-analyses seem to show that the programs for probationers or juvenile offenders with community aftercare are inclined to be more effective. There have been major theoretical and methodological advances in the juvenile and adult correctional treatment literature since Martinson's assessment study. Expecially, adaptation of the psychological learning model and meta-analysis as a statistical technique to criminology and corrections have been known to contribute in formulating principles for successful treatment programs though those principles need further clarification and empirical assessment. The following are suggested as principles of successful correctional treatment: linkage with criminogenic needs, multimodal programs for various deficits, matching client learning styles with staff teaching styles, treatment based on risk differentiation, providing skills oriented to cognitive-behavioral treatment, implementing programs with continuity of care and sufficient dosage, and involvement in both program development and evaluation by researchers.