Despite the overall decline in cigarette smoking prevalence in the US, social class inequalities in smoking are likely to persist, or even to widen. One possible reason for the increasing gap in smoking prevalence across social class could be our lack of understanding of causal mechanisms: in other words, what accounts for the social gradient in smoking behavior? In this paper, we examine the mechanisms behind social gradients related to smoking cessation by use of path analysis techniques. The data come from a 3-year follow-up telephone survey of a cohort of US adults. The sample for the present analysis was drawn from the 481 respondents who reported being smokers and employed at baseline and who completed the follow-up interview. We examined two social class indicators, educational attainment and household income, in relation to smoking cessation. We tested the potential mediating effects of the following variables: differential use of resources for smoking cessation (e.g., booklet, pamphlet, quit line, nicotine replacement therapy and smoking cessation program), differential environments in terms of smoking at worksite and home, and differences in peer smoking. Our path analyses suggest that smokers from high social class are likely to use effective resources for smoking cessation and have restrictive home environment in terms of smoking, which leads to a relatively higher smoking cessation rate compared to those from low social class. The results of this study suggest that interventions should target resources for smoking cessation and home environments in terms of smoking to reduce socio-economic disparities in smoking cessation.