The question of whether punishments effectively achieve their intended purpose is the central premise of the criminal justice system and is deceptively straightforward. This study analyzes the three major justifications for punishing offenders: deterrence, incapacitation and rehabilitation, against empirical evidence found in both common law and civil law system sentencing practices. Significant criminological research indicates that the credibility of deterrent sentencing has been exaggerated. The rate of return benefits from incapacitation decreases rapidly as the imprisonment population grows larger (i.e., after a certain point, the rest of the population is not at risk from imprisonment). Finally, empirical rehabilitation programmes have been successfully shown to provide the best option for reducing recidivism and below is an analysis of three commonly discussed sentencing policy models – mandatory minimums, truth-in-sentencing, and restorative justice; and this paper concludes by proposing a graduated framework, which supersedes the socio-political influences of penalty populism by focusing on proportionality and empirical outcomes.