This study addressed several issues related to the course of smoking among young women entering the US Navy. The smoking experience of women entering the Navy was described, and their smoking prevalence was compared to civilians. The effect of an involuntary, 8-week smoking ban during recruit training on recruits' smoking perceptions and intentions was examined. Three months after leaving recruit training, baseline smokers were reassessed to determine the impact of the eight weeks of enforced cessation on smoking relapse. Finally, nonresponse to the 3-month assessment and its potential effect on validity of findings was investigated. About 5,500 women were surveyed over a 1-year period at entry to recruit training. Standardized comparisons showed the smoking rate among recruits was higher than civilians' (38.7% versus 28.8%, respectively) controlling for differences in age, race, and education. Age, education, and race/ethnicity were related to smoking prevalence at entry. As a result of the 8-week smoking ban, there was a 39% reduction in the percent of all women recruits who perceived themselves as smokers from entry to graduation from recruit training. Age, education, race/ethnicity, level of baseline smoking and addiction, and intentions to smoke were related to changes in perceptions of being a smoker, or the maintenance of consistent positive perceptions of being a nonsmoker. Similarly, changes in intentions to smoke were associated with sociodemographic and baseline smoking variables. Those with any smoking experience at entry were surveyed 3-months after leaving recruit training to determine subsequent short-term effects of the involuntary smoking ban. Response rate to the follow-up survey was 39%. Reluctant respondents were not different from those responding promptly to the original survey, but nonrespondents as a group were less educated, had a higher prevalence of baseline smoking, and were more frequent smokers that respondents. Sixty-eight percent had relapsed at the follow-up. Relapse rates varied by sociodemographic characteristics, entry and graduation smoking variables, entry-to-graduation changes in perceptions and intentions to smoke, and Navy factors. The type of smoker one was at entry to the Navy was an important predictor of relapse, with 89% of daily smokers smoking at the 3-month follow-up. Results suggest that the 8-week involuntary ban on smoking was useful in bringing about changes in self-reported smoking status and in helping some individuals stay off cigarettes once they leave recruit training, although policy alone cannot be expected to be effective for all smokers. High relapse rates particularly among heavier smokers underscore the need for educational/motivational efforts during recruit training to help maintain the quit status achieved during those eight weeks.