The efficacy of cimetidine as a treatment that could reduce smoking in heavily dependent smokers has been determined. In a randomised, double-blind, double-crossover experiment, 43 heavy smokers were divided into two groups, one receiving cimetidine 400 mg orally three times a day, and the other receiving placebo for two weeks followed by the alternative treatment (placebo or cimetidine). No significant difference in the mean alveolar carbon monoxide, nicotine or cotinine levels was found between the two treatment groups compared to baseline. Since the alveolar carbon monoxide level reflects the intensity of smoking behaviour, the results suggest that no change in smoking behaviour occurred in the subjects. Contrary to our previous findings that cimetidine decreased the total body clearance of nicotine by 30% in a population of non-smokers, in the heavily dependent smokers, cimetidine did not appear to alter nicotine elimination. One possible explanation for the discrepancy is that tobacco smoking is known to induce nicotine metabolism and the induction might have offset any effect of cimetidine on nicotine elimination. Cimetidine does not appear to be a useful treatment leading to a reduction or cessation of cigarette smoking.