The role of active and passive cigarette smoking in breast cancer etiology remains controversial. Using data from a large population-based case-control study in Poland (2386 cases, 2502 controls) conducted during 2000-2003, we examined the associations between active and passive smoking overall and for different age categories. We also evaluated differences in risk by estrogen receptor (ER) and progesterone receptor (PR) status in tumors, and the potential modification of the smoking association by N-acetyl transferase 2 (NAT2) genotype. Women ever exposed to passive smoking at home or at work had a risk of breast cancer similar to those never exposed to active or passive smoking (OR (95%CI) 5 1.11 (0.85-1.46), and no trends were observed with increasing hours/day-years of passive smoking exposure. Active smoking was associated with a significant increase in risk only among women younger than 45 years of age (OR (95%CI) 5 1.95 (1.38-2.76); 1.15 (0.93-1.40); 0.91 (0.77-1.09) for <45, 45-55 and >55 years of age, respectively; p-heterogeneity <0.001 for <45 vs. >55 years) and prevailed for both ER1 and ER2 tumors. The smoking association among women <45 years was stronger for current than former smokers, and a significant trend was observed with duration of smoking (p 5 0.04). NAT2 slow vs. rapid/intermediate acetylation genotype was not related to breast cancer risk (0.99 (0.87-1.13)), and did not significantly modify the smoking relationships. In conclusion, our data indicate that passive smoking is not associated with breast cancer risk; however, active smoking might be associated with an increased risk for early onset breast cancers. ' 2006 Wiley-Liss, Inc.