ABSTRACT Aim To compare the characteristics and outcomes of users accessing pharmacy and group-based smoking treatment. Design Observational study of administrative information linked with survey data. Setting Glasgow, Scotland. Participants A total of 1785 service users who set a quit date between March and May 2007. Intervention Smoking treatment services based in pharmacies providing one-to-one support, and in the community offering group support. Measurements Routine monitoring data included information about basic demographic characteristics, deprivation category of residence, nature of intervention and smoking status at 4-week follow-up determined by carbon monoxide (CO) readings ≤10. These data were supplemented by information about socio-economic status and smoking-related behaviours obtained from consenting service recipients by treatment advisers. Findings In the pharmacy-based service 18.6 % of users (n = 1374) were CO-validated as a quitter at 4 weeks, compared with 35.5 % (n = 411) in the group-based service. In a multivariate model, restricted to participants (n = 1366) with data allowing adjustment for socio-demographic and behavioural characteristics and including interaction terms, users who accessed the group-based services were almost twice as likely (odds ratio 1.980; confidence interval 1.50–2.62) as those who used pharmacy-based support to have quit smoking at 4-week follow-up. Conclusions Specialist-led group-based services appear to have higher quit rates than one-to-one services provided by pharmacies but the pharmacy services treat many more smokers. More research is needed to determine what can be done to bring the success rates of pharmacy services up to those of specialist-led groups and how to expand access to group-based services.