OBJECTIVES A variety of pharmaceutical treatments for smoking cessation are now available. We conduct a systematic review on economic evaluations of these therapies between 2000- 2007, ranked them by cost and effectiveness, and showed study limitations in the process of defining efficient public policies. METHODS A systematic search was conducted. PubMed (U.S. National Library of Medicine) and the NHS Economic Evaluation Database from the Centre for Reviews and Dissemination were reviewed in detail using standard criteria. RESULTS Out of 870 relevant publications, 49 complied with the inclusion criteria. The documents were very heterogeneous in terms of the following elements: perspective, methodology and effectiveness measures used; included costs; evaluated interventions, and studied population. The cost per life year gained, adjusted or not for quality, ranged between US$367 and US$7284 (December 2006). In most of them, bupropion (with or without medical advice) was the most cost-effective option. Its incremental cost per life year gained, adjusted or not for quality, ranged from US$454 to US$2984 (December 2006). However, recent articles compared varenicline with bupropion and the nicotine patch and showed varenicline to be the most cost-effective. Nevertheless, additional evaluations of varenicline, including relapse rates over longer periods of time, are needed. CONCLUSIONS Useful literature regarding the economic implications and health benefits of pharmacotherapy for smoking cessation is available. The results suggest that pharmacotherapy could be cost-effective in middle-income countries. Nevertheless, the lack of cost-effectiveness data from low- and middle-income countries cries out for more analyses that deal with the way in which the estimates change when models attempt to reproduce poorer settings. In addition, to obtain more robust estimates, it will be necessary to develop models that better predict the success of pharmacotherapy, incorporating adherence to therapy under non-controlled conditions, and the relationship between the smokers' behavior and the probability of cessation.