Aims To examine the impact of socio-demographic factors, smoking-related behaviour and service characteristics on CO-validated quit rates at 4-week follow-up in English smoking treatment services, and to compare the results with those for self-reported quitters. Design Observational study of administrative information linked with survey data for 6959 recipients of smoking treatment services who set a quit date between October 2001 and March 2003. 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-week follow-up. These data were supplemented with additional information about referral pathways, socio-economic status and smoking-related behaviours obtained from consenting service recipients by NHS advisers. Findings More than one-half of clients (53%) were CO-validated as quitters at 4 weeks, rising to 60.7% when self-reported cases were included. Age (OR 1.026; Cl 1.022-1.029) and being extremely determined to quit (OR 1.46; CI 1.26-1.71) were associated positively with CO-validated cessation, whereas women (OR 0.85; Cl 0.77-0.94), users with lower socio-economic status (OR 0.92; CI 0.88-0.95), those smoking 31 or more cigarettes daily (OR 0.75; Cl 0.64-0.88) and those with relatively poor health status (OR 0.72; CI 0.63-0.82) were less likely to quit. Although the vast majority of users received one-to-one support, those who had group counselling were more likely to be successful in their quit attempt (OR 1.38; Cl 1.09-1.76). Self-report and CO-validated quitters were similar in terms of their characteristics. Conclusions These results obtained from routine services support those obtained from clinical trials and confirm the effectiveness of counselling combined with pharmacotherapies to assist smokers to quit in the short term. However, the relative effectiveness of group interventions raises questions about why one-to-one counselling is used much more commonly. The importance of socio-demographic and nicotine-related dependency factors also suggests that local service targets for smoking cessation need to take account of the social distribution of these characteristics.