Evidence suggests that place of residence may be associated with smoking independently of individual poverty and socio-economic status. Qualitative research undertaken in disadvantaged communities in Glasgow explored possible pathways which might explain this 'area effect'. A poorly resourced and stressful environment, strong community norms, isolation from wider social norms, and limited opportunities for respite and recreation appear to combine not only to foster smoking but also to discourage or undermine cessation. Even the more positive aspects of life, such as support networks and identity, seem to encourage rather than challenge smoking. Policy and intervention responses need to tackle not only individual but also environmental disadvantage.