Smoking cessation interventions in the UK are being developed in the context of widening socio-economic differentials in both prevalence and cessation. These differ entials are evident among pregnant women, the group targeted for interventions directed at women. Recent research has suggested that, among the disadvantages associated with low socio-economic status, being dependent on means-tested benefits may be a particularly powerful influence on snuokimg status and a major barrier to quitting. Intervention programmes have been heavily influenced by the transtheoretical model, which maps the quitting process as a patterned sequeuce of 'stages of change'. However, little is known about the stage-of-change profile in the UK population or about the socio- economic patterning of the profile. This paper begins to fill these gaps in the knowledge base of health promotion with respect to women in pregnancy. It reports on a survey of 2000 expectant mothers conducted in 1996 in the West Midlands. First pregnancy was found to have an intervention-like effect, with a high proportion of first-time expectant mothers who entered pregnancy as smokers either planning to quit or having done so. This intervention-like effect was moderated by women's socio-economic circumstances. Being in receipt of means-tested benefits increased the odds of a woman not intending to give up smoking in the foreseeable future.