Aims. Nicotine replacement therapy (NRT) is an established aid in stopping smoking, while the role of antidepressants remains uncertain. Antidepressants added to NRT might improve abstinence rates. Our aim was to determine the efficacy of nicotine inhaler and fluoxetine vs. nicotine inhaler and placebo in attempts to quit smoking. Design. A randomized, double-blind, placebo-controlled trial. Setting. A smoker's cessation clinic. Participants. One hundred volunteers smoking 10 cigarettes/day or more. Interventions. Subjects were instructed to start taking a daily dose of 10 mg of fluoxetine or placebo 16 days before stopping smoking, then 20 mg 10 days before quitting, continuing for up to at least 3 months. Subjects were instructed to use 6-12 units per day of nicotine inhalers after stopping smoking for up to 6 months. Measurements. Continuous abstinence rates recorded at various time points up to 12 months from the quit date. Findings. The sustained abstinence rate for the inhaler-fluoxetine group was 54%, 40%, 29% and 21% after 1.5, 3, 6 and 12 months, respectively, compared to 48%, 40%, 32% and 23% for the inhaler-placebo group. The differences were not significant at any time point. Abstinence up to 3 months was more likely in older smokers, those with a lower Beck Depression Inventory Score (BDI), lower Fagerstrom Test of Nicotine Dependence (FTND) score and no history of alcoholism. Fluoxetine appeared to increase abstinence rates among high BDI smokers compared to high BDI smokers assigned placebo. Serum levels of nicotine during treatment in the inhaler-fluoxetine group were lower than in the inhaler-placebo group so that fluoxetine may have reduced inhaler use through a common site of action. Conclusions. We found no evidence that fluoxetine treatment when used as an adjunct to NRT in unselected smokers is effective, but there may be an advantage to using it in depressed smokers.