Introduction: general practice is a recommended setting for the delivery of smoking cessation programs. Little is known about the types of practice that achieve higher cessation rates. To address this gap in knowledge, we assessed the impact of general practice characteristics on the outcomes of a large scale smoking cessation intervention delivered in general practice settings. Method: a cross-sectional study was undertaken of 7,778 participants enrolled on a structured cessation program comprising repeated brief interventions in one-to-one sessions and nicotine replacement therapy in Christchurch New Zealand, 2001–2007. We employed a logistic multilevel analysis of respondents nested in general practices with cessation at 6 months as the outcome measure. Results: after taking into account relevant individual-level predictors (age, sex, smoking intensity) and area-level surrogates for individual predictors (socioeconomic status and access to tobacco retail outlets), there remained significant variation in quit rates between practices. This variation reduced when practice characteristics were included. Practices with a majority of male doctors and practices with fewer male patients were associated with better quit rates. Practices with large numbers of doctors were less effective in achieving cessation with heavy smokers. Conclusions: the effectiveness of smoking cessation programs can be influenced significantly by practice characteristics. To increase quit rates, more attention should be paid to the institutional setting of smoking cessation programs. Assessments of the effectiveness of cessation programs should give appropriate recognition to the fact that some practices may find higher quit rates more difficult to achieve