Objective: To determine which factors predict default from subsequent treatment sessions after initial enrolment and attendance at a large, English smoking cessation service. Design: Cross-sectional survey using data obtained at smokers' initial enrolment attendance to compare the characteristics of those who subsequently default with those who do not. Participants and setting: 2516 smokers who attended at least one appointment with a smoking cessation service serving the Nottingham City area, UK, in a two year period starting from 1 October 2001. Results: The 2516 smokers who consented for their data to be used in this analysis comprised 78.8 per cent of all smokers enrolled into the smoking cessation service during the study period. Of these, 73.9 per cent (1892) attended subsequent appointments, whilst 21.6 per cent defaulted (i.e. failed to re-attend for treatment after the initial appointment). Default from treatment was independently associated with younger age, heavier smoking, living alone or in a house with a smoker, and being less motivated to stop smoking, but was unrelated to smokers' routes of referral to the service. Conclusions: Further work is required to explain why described factors are associated with default from smoking cessation treatment in this service and to determine whether similar factors influence this in other services. Our results suggest that, to minimize default, smoking cessation services should, at initial enrolment consultations, employ strategies to increase smokers' motivation to stop, combat nicotine addiction and the influence of other smokers at home. (Contains 3 tables.)