Introduction: More than 2 million persons are incarcerated in the United States. Most are young minority men, soon to reenter the community. The majority are also lifelong smokers with high rates of health-related problems. As prisons implement smoking bans, it is not known whether health behavior change that is mandated, rather than selected, can be maintained. The Wisconsin Department of Corrections smoking ban is a unique opportunity to investigate determinants of smoking behavior after release from prison. Methods: A convenience sample of 49 incarcerated men near release participated in two interviews (1-month prerelease, in prison, and 1-month postrelease via telephone). Descriptive analyses and multivariate modeling were conducted to determine associations with postrelease smoking. Results: Participants had a mean age of 36.7 years, 12.4 years of education, and a 2.3-year incarceration; 47% were Black and 41% White. They had smoked 14.5 years. Most (67%) believed that their health was improved after the smoking ban. Paired t tests revealed decreases in Positive and Negative Affect Scale negative affect (p = .001) and Patient Health Questionnaire-8 depression (p = .009) postrelease. Univariate analysis showed correlations of intent to smoke upon release with smoking relapse postrelease (p = .001), White race with smoking relapse postrelease (p = .045), and perceived better health since the prison smoking ban with nonsmoking on release (p = .01). There was a trend toward use of alcohol with smoking relapse on release (p = .061). Discussion: Prerelease smoking intention predicted postrelease behavior. Belief in improved health after the prison smoking ban correlated with nonsmoking on release. Targeted relapse prevention interventions are needed for people reentering the community.