The United States and other industrialized nations have largely based their criminal justice systems on the deterrence model, which states that punishing offenders for their crimes deters them from committing future offenses. This deterrence theory has been one of the cornerstones of modern criminological and criminal justice thought. However, surprisingly, recent studies have suggested that offenders who are punished for their crimes are likely to re-offend because they believe they are unlikely to be punished for subsequent crimes. The authors examine two explanations for this finding. First, they explore the explanation called “selection,” which is consistent with deterrence theory. The selection explanation holds that those offenders most likely to be arrested are the most active and prolific offenders. It logically follows that these active offenders are likely to believe they are impervious to the law. The second explanation is called “resetting” and holds that offenders who have been arrested for their crimes believe that they would have to be very unlucky to be arrested a second time. The authors then report on a preliminary study that examined both of these explanations. The results did not positively identify either of these theories as explanations for offenders’ belief in the uncertainty of future punishment. Future research should investigate the possibility of alternative explanations. Tables, notes, references