FMHA-based research and practice is concerned with the measurement of any psychologically relevant construct that contributes to decisionmaking in a legal context. Although researchers justifiably continue to improve risk assessment for violence, empirical investigations of FMHA for nonviolent offenders are rare. This is surprising and unwarranted for several reasons. First, nonviolent offenders compose the majority of individuals incarcerated in U.S. Federal and State prisons and local jails as of 2003. Much of the overcrowding crisis in jails and prisons has been attributed to the rising number of nonviolent offenders being incarcerated. Second, the costs of such incarceration coupled with high rates of recidivism among nonviolent offenders after their release suggests that the current sentencing of nonviolent offenders is not cost-effective. The few FMHA-based studies with nonviolent offenders show that they generally pose less risk for future violence and less risk for general recidivism, manifest fewer severe problems across a range of domains, and may have less personality pathology than violent offenders. Given the recent trend toward "evidence-based corrections" and alternatives-to-incarceration approaches for nonviolent offenders, many are recognizing the importance of FMHA in accounting for static and dynamic risk factors and intervention needs of nonviolent offenders. Empirical studies of FMHA with nonviolent offenders show it can distinguish low-risk nonviolent offenders from moderate-risk and high-risk nonviolent offenders, such that accurate predictions of postrelease outcomes and differential performance in community-based programs are possible. The article notes current limitations in research on FMHA nonviolent offenders and offers suggestions for future research. 59 references