Although smoking prevalence among pregnant women is highest among those of lower socio-economic status, the factors associated with cessation in this high risk group are poorly documented. This paper reports on data from a survey of working class women in Nottingham and Coventry, who were interviewed 6 months after having delivered a baby. It compares the characteristics of the 32 women who quit smoking during their pregnancy and were still ex-smokers at the 6 month post-natal point, with 472 women who had continued to smoke during their pregnancy or had returned to smoking in the post-partum period. A logistic regression analysis revealed three variables that were independently associated with long term cessation-having previously quit for more than 1 week, having a non-smoking partner, and believing that the children of smokers are more likely to get infections. Among the most disadvantaged, an understanding of how close social ties and economic circumstances influence smoking behaviour is crucial, if we are to develop more effective smoking cessation intervention strategies for working class women.