Aims To assess the impact of English treatment services on CO-validated quit rates at 52-week follow-up, to explore the relationship between service-related characteristics and socio-demographic and behavioural factors with cessation outcomes, and to compare the characteristics of service users lost to follow-tip with CO-validated quitters. Design Observational study of administrative information linked with survey data for 2069 recipients of smoking treatment services who set a quit date between May and November 2002. Setting Two contrasting areas of England, Nottingham and North Cumbria, consisting of nine primary care trust (PCT) localities. Measurements Routine monitoring data specified by the Department of Health included information about basic demographic characteristics, postcode of residence from which a deprivation category was identified, nature of intervention, and smoking status at 4-weck follow-up. These data were supplemented with information about smoking status at 52 weeks, referral pathways, relapse experiences, number of follow-up contact attempts, socio-economic status and smoking-related behaviours obtained from consenting service recipients by treatment advisers. Findings One user in seven (14.6%,) reported prolonged abstinence and was CO-validated as a successful quitter at 52 weeks. This rose to 17.71% when self-report cases were included. Relapse rates between 4 and 52 weeks were almost identical between the two study areas-75%). Relapse was most likely to occur in the first 6 months following treatment. Users who self-reported quitting at 4 weeks were less likely (13.7%) than those with biochemical verification of smoking status at 4 weeks (2 5.2 %) to be CO-validated quitters at 5 2 weeks (P=0.004). Older users (OR 1.023; Cl 1.014-1.032), people who smoke mainly for pleasure rather than to cope (OR 1.38; Cl 1.02-1.8 7), and those who were extremely determined (OR 1.58; CI 1.21-2.05) were more likely to be quitters at 52-week follow-up. whereas those with lower socio-economic status (OR 0.86; Cl 0.78-0.96), who smoked their first cigarette of the day within 5 minutes of waking (OR 0. 7 3; Cl 0. 5 5-0.9 6) or had another smoker in their household (OR 0. 6 5; Cl 0. 49-0.8 6) were less likely. In contrast, users lost to follow-up tended to be younger and experienced different referral pathways than CO-validated quitters. Gender was not statistically significantly associated with cessation at 52 weeks and nor were any of the key characteristics of intervention, such as group or one-to-one counselling. Conclusions These results obtained from routine services are consistent with those obtained from clinical trials in relation to abstinence at one year. Given that a high proportion of smokers relapsed between 4 weeks and I year it is important that future assessments of longer-term outcomes are conducted. However, following-up service users many months after an intervention is expensive, and reasonable estimates of quit rates can be estimated from short-term outcomes, provided that they have been CO-validated. Future studies should monitor outcomes from a selection of services treating different groups of smokers, particularly if more is to be learned about the role of smoking treatment services in reducing inequalities in health.