Noting that alarming, worldwide increases in crime and recidivism rates have occurred in the 200 years following the abandonment of punishment as a penological philosophy in favor of treatment-oriented programs for convicted offenders, some criminologists have fallen into the fallacy of post-hoc reasoning by blaming progressive correctional policies for current crime problems. Even some liberal intellectuals in the United States are advocating a general return to fixed and equitable, but punitive sentences for offenders, based solely on the provisions of the criminal code. Some States have already started abolishing parole programs in keeping with the new regressive penological trend, contributing to the reemergence of the treatment model versus justice model controversy. Etiological research into the causes of crime as a basis for the individual treatment of offenders is in danger of being neglected, even abandoned, despite its contributions to the identification of criminogenic factors to explain how crime is generated. Inroads by biocriminology, with emphasis on biological and genetic, i.e., hereditary, versus cultural and sociological criminogenic factors, have added to the claims of the return-to-punishment advocates. Nevertheless, United States penologists are still trying to point to favorable results of the treatment-model by collecting further data and using the latest evaluation techniques. There is still hope, therefore, that the role of etiology in criminological research will be upgraded, rather than downgraded. This is especially vital in view of the necessity to clarify the new transnational dimensions of crime and violence in the aftermath of wars, natural disasters and genocide attempts which have caused large-scale migrations and the criminogenic uprooting of millions of people. A rational approach to penology appears to be a composite of the treatment and justice models, with all the variations needed to meet individual situations, the effectiveness of which should be evaluated on a continuing basis.