The high rates of offenders returning to prison after their release have renewed interest in explanations of recidivism. While researchers have focused on individual level crime theory and characteristics to explain recidivism, one neglected area of study is neighborhood context’s role in recidivism. The few examples of research that attempts to address this gap also fail to incorporate the fact that recently released offenders are likely a highly mobile population, something that could cause error in contextual analysis.Using a unique dataset of released offenders in Ohio that contains serial residence and violation information, this project tests a social disorganization model of recidivism, while also analyzing mobility patterns. A third analysis uses a unique property of the dataset: the fact that it includes prisoners sentenced under indeterminate and determinate sentencing guidelines. This offers the chance to assess the deterrent aspects of the switch in sentencing as well as how deterrence interacts within neighborhood context.The first analysis of residential mobility patterns answers four questions: do recently released offenders move, where do recently released offenders move, who lives where initially, and who moves? The results indicate that offenders do move and move at a higher rate than the general public, with most offenders in the study experiencing at least one move in the first year after release. These movements are mostly lateral, or to neighborhoods that are similar to the neighborhoods offenders reside in immediately after release. Neighborhood of initial residence is related most notably to race and whether an offender is released into a halfway house. Finally, movement is related to employment, race, and living in a halfway house.The second analysis of contextual effects on recidivism, using discrete time multilevel logistic regression models to predict likelihood of a parole violation and arrest, reveals support for social disorganization theory. The results show neighborhood economic disadvantage and residential stability are related to individual offender’s likelihood of recidivism. Individual characteristics, like demographic characteristics, prior offending history, employment and a time variant measure of residential mobility are also related to recidivism.The third analysis uses a quasi-experimental method to test deterrence between parolees and PRC offenders. After propensity score matching a sample of both groups, the results reveal little evidence of a deterrent effect of length of potential punishment on recidivism beyond a consistent parolee/PRC effect. Interestingly, contextual measures only mattered for parolees, indicating that deterrence and context may interact.The results from this set of analyses indicate three major findings. First, while prisoners move, that movement is mostly lateral, meaning that contextual research that has not taken movement into account is likely unbiased. Second, neighborhood context does matter for recently released offenders, as offenders in economically disadvantaged neighborhoods and less residentially stable neighborhoods are more likely to recidivate. Third, recently released offenders showed little evidence of being deterred, although there was some evidence that parolees in disorganized neighborhoods were more likely to recidivate than parolees in better neighborhoods, indicating a potential undermining of deterrence in disorganized neighborhoods.