Our current criminal system generally accepts the notion that a jury ought to convict on the basis of the facts adduced at trial, not out of a belief that the defendant’s character was “bad” so certainly he acted in conformity with that character in committing the present offense. Despite this understanding, former convictions may be entered into evidence under certain circumstances. As a result, evidence of past criminal activity may be used to fashion a convicted defendant’s sentence for a present offense. This article evaluates competing definitions of what constitutes a federal “first time” offender and identifies options for incorporating a first offender provision into the federal sentencing guidelines. It first examines the status of recent research defining a federal first time offender, and describes the data and methods used in the analysis. It then contrasts the characteristics of the three proposed first offender groupings among themselves, and in relation to offenders in the remaining criminal history categories, to demonstrate the unique nature of federal offenders with little or no criminal history. The article looks at the impact of various first offender definitions on the predictive power of the guidelines, and then proposes first offender provision options and discusses how first offender sentencing alternatives might be implemented. The final section summarizes the findings of the study, highlighting policy and empirical debates that require resolution prior to implementation, including specific legal conflicts rising from the introduction and operationalization of a first offender provision. Essentially, the article concludes that the criminal history categories used in the federal sentencing guidelines have served as a reasonably reliable indicator in determining both culpability and the likelihood the offender will commit future criminal acts. The article does suggest, however, that the criminal history categories need to be refined to improve recidivism measures.