This dissertation examines whether solitary confinement and time spent in solitary confinement impacts institutional conduct upon release from solitary confinement. This report will also help correctional professionals and policy makers make well-informed decisions about the use of solitary confinement. "Solitary confinement (SC) has been an important component of the American prison system since the emergence of the penitentiaries in the early 1800s. The main criticism of SC has long been that it causes inhabitants undue psychological distress and by extension increases propensity toward criminal behavior. The use of SC raises constitutional and humanitarian concerns, with critics who charge the practice constitutes cruel and unusual punishment, is inhumane, and violates the minimum standards of decency. However, SC is also a management tool in which correctional officials have come to rely upon for the effective management of prisons, and many would not waiver in the contention that SC is needed to ensure the safety and security of these institutions" (p. i). This study examines what effect solitary confinement has on the futures criminal behavior of offenders released from SC. The impacts of SC on mentally ill inmates, gang members, gender, risk, race, age, offense type, sentence length, and custody level are presented. "The most important finding in this study is the lack of evidence of any effect of SC on subsequent inmate misconduct. In all twelve of the multivariate models examined here, SC was not significantly related to misconduct. These results suggest that neither the experience of SC, nor the number of days spent in SC, had any effect on the prevalence or incidence of the finding of guilt for subsequent violent, nonviolent, or drug misconduct. These findings run counter to the arguments that SC decreases, or increases, criminal behavior and support the conclusion that SC has no effect on criminal behavior. Further, seriously mentally ill inmates in SC had an increased risk for subsequent nonviolent and drug misconduct, while gang members in SC had an increased risk for subsequent violent and nonviolent misconduct. This study did not reveal much of a difference in effect based on gender, though there is some evidence that females in SC may be less likely than males to engage in subsequent violent misconduct. Risk was not found to have any significant relationship. However, it is cautioned that the risk assessment used was less than ideal and further research should be conducted with other risk instruments before any definitive conclusion are made about the mediating effect of risk on institutional behavior. There were no differences found in the effect of SC based on race or prior incarceration. Younger inmates in SC were found to be at a increased risk for violent and nonviolent misconduct. Finally, sentence offense type did not have much of an influence on misconduct, though being committed for a drug offense showed a reduced risk in violent misconduct when compared to nonviolent offense" (p. 112-113).