Smoking prevalence across high-income countries such as the United Kingdom has significantly decreased over the past few decades; this decrease, however, has not occurred uniformly across social strata. The highest concentrations of smokers are currently found in lower-income groups. Lack of access to material resources and differing social norms have been cited as possible causes of this imbalance in smoking behaviour. Social capital, measured by trust and levels of community participation, has also been postulated to influence health behaviour. Data from the British Household Panel Survey were used to identify smoking and non-smoking cohorts at baseline (N = 10,512); from these, individuals whose smoking behaviour had changed (the dependent variable) were identified. Measures of social capital, income, employment and marital status, and considered confounders were tested for associations with changes in smoking behaviour over a 2-year period. Both bivariate and multivariate models were utilized to elicit associations. Only marital and employment status, along with social capital measures, remained significantly associated with changes in smoking behaviour. Individual/household income, baseline social class and general/psychological health failed to demonstrate any significant association with changes in smoking status. Support mechanisms (via marriage and employment) and elements social capital (measured by 'trust' and 'social participation') are independently and positively associated with smoking cessation; continual lack of active social participation and remaining single are associated with smoking initiation. Smoking interventions should consider increased participation as an intrinsic part of their design.