OBJECTIVES: In healthy smokers, antidepressants can double the odds of cessation. Because of its four times lower costs and comparable efficacy in healthy smokers, nortriptyline appears to be favourable compared to bupropion. We assessed which of both drugs was most effective and cost-effective in stopping smoking after 1 year compared with placebo among smokers at risk or with existing chronic obstructive pulmonary disease (COPD). METHODS: A total of 255 participants, aged 30-70 years, received smoking cessation counselling and were assigned bupropion, nortriptyline or placebo randomly for 12 weeks. Prolonged abstinence from smoking was defined as a participant's report of no cigarettes from week 4 to week 52, validated by urinary cotinine. Costs were calculated using a societal perspective and uncertainty was assessed using the bootstrap method. RESULTS: The prolonged abstinence rate was 20.9% with bupropion, 20.0% with nortriptyline and 13.5% with placebo. The differences between bupropion and placebo [relative risk (RR) = 1.6; 95% confidence interval (CI) 0.8-3.0] and between nortriptyline and placebo (RR = 1.5; 95% CI 0.8-2.9) were not significant. Severity of airway obstruction did not influence abstinence significantly. Societal costs were 1368 euros (2.5th-97.5th percentile 193-5260) with bupropion, 1906 euros (2.5th-97.5th 120-17 761) with nortriptyline and 1212 euros (2.5th-97.5th 96-6602) with placebo. Were society willing to pay more than 2000 euros for a quitter, bupropion was most likely to be cost-effective. CONCLUSIONS: Bupropion and nortriptyline seem to be equally effective, but bupropion appears to be more cost-effective when compared to placebo and nortriptyline. This impression holds using only health care costs. As the cost-effectiveness analyses concern some uncertainties, the results should be interpreted with care and future studies are needed to replicate the findings.