Detailed individual level longitudinal data on addictive behaviors such as smoking, consumption of alcohlic beverages, and illicit drug use are largely nonexistent. This has made separating the determinants of addictive behavior and cessation from unobserved individual characteristics difficult at best for researchers. This paper uses the first four waves of U.S. data from the International Tobacco Control Survey to examine smoking cessation and cigarette tax avoidance behaviors. In addition, the paper addresses the potential introduction of bias towards smokers that are least likely to quit or alter their behaviors through survey attrition.