The association of breast cancer with passive and active smoking was investigated in slow and fast acetylators of aromatic amines in a Geneva, Switzerland, study in 1996-1997. A slow acetylator was homozygous for one, or heterozygous for two, of three N-acetyltransferase 2 (NAT2) polymorphisms determined on buccal cell DNA from 177 breast cancer cases and 170 age-matched, population controls. The reference group consisted of women never regularly exposed to active or passive smoke. Among premenopausal women, the odds ratios were homogeneous in slow and fast acetylators: 3.2 (95% confidence interval (CI): 1.2, 8.7) for passive smoking and 2.9 (95% CI: 1.1, 7.5) for active smoking. Among postmenopausal women, the odds ratios for fast acetylators were 11.6 (95% CI: 2.2, 62.2) for passive and 8.2 (95% CI: 1.4, 46.0) for active smoking; the corresponding effects were also apparent but less strong in slow acetylators. After the nonexposed and the passive smokers were grouped in a single reference category, active smoking was associated with postmenopausal breast cancer in slow acetylators (odds ratio (OR) = 2.5, 95% CI: 1.0, 6.2) but not in fast acetylators (OR = 1.3, 95% CI: 0.5, 3.3). Thus, the associations of both passive and active smoking with breast cancer appear stronger in fast than in slow NAT2 genotypes. Separating passive smokers from the nonexposed impacts on the inference about a possible NAT2-smoking interaction.