A review of Robert Martinson's "nothing works" sentencing doctrine indicates two major flaws: (1) it relies on recidivism as the sole measure of sentencing success; and (2) it fails to address the issue of how sentences are implemented and operate in practice. Although recidivism has traditionally been the criterion used to measure the success of sentences, there is a real issue as to whether it is the right criterion and whether it should apply to all sentences. Certainly, a major goal of probation-based disposals is to reduce reoffending; whether this is the case for prisons and fines is debatable. The author proposes that multiple measures, including both primary and secondary measures, be used to evaluate sentence effectiveness. 19 references