OBJECTIVE: To review how health economic models of the cost-effectiveness of smoking cessation interventions have evolved over the last ten years. METHODS: The Medline database was searched via PubMed using the keywords ‘cost-effectiveness models smoking cessation'. Articles not available or dating before 1998 were excluded. The abstracts from these results were examined, and relevant studies were selected for the review. A bibliographic search was implemented on the selected studies and additional relevant studies were included. An extraction grid was built to record key comparable attributes of each included model as reported. RESULTS: Thirteen studies were included and extracted, dating from 1999-2008. The most common model design was Markov model simulations of hypothetical patient cohorts given various interventions (6 studies). Reported cost perspectives have widened over the period: since 2004, societal perspectives have been the most common (7/8 studies) whereas previously direct perspectives were more common (4/5 studies). Prior to 2004, 3/5 studies were of UK patients whereas since then 4 were for the US, 3 Scandinavia. The interventions included in the models were initially Nicotine Replacement Therapy (NRT) and bupropion (Zyban), but this has widened recently to also include varenicline (Champix), which was approved by the FDA in May 2006. It was not possible to draw clear conclusions as to the values of cost, health outcome and cost-effectiveness made given the wide variety of study designs, perspectives and countries modelled and the small number of studies. CONCLUSION: There is wide variety in the nature of published health economic models of smoking cessation in terms of perspective, country and model structure. Some trends have been observed. Markov models were the most common model design to use for modelling smoking cessation cost-effectiveness in the articles we reviewed.