Although the involvement of environmental tobacco smoke (ETS) in human lung cancer is no longer a matter of dispute, the magnitude of its impact still is. This is mainly due to the inefficiency of methodology to assess exposure to ETS especially in public places. Setting a real life exposure condition (3 h stay in local pubs) and using a matched-control study design, we quantified smoke-related DNA adducts in induced sputum and peripheral blood lymphocytes (PBL) of healthy non-smokers (n = 15) before and after a single pub visit by means of the (32)P-post-labeling assay. For verification, we also measured a spectrum of polycyclic aromatic hydrocarbons (PAH) in the ambient air of the pubs by personal air monitors, and determined the plasma concentrations of nicotine and cotinine by gas chromatography/mass spectrometry. The ambient air concentrations of all PAH were several orders of magnitude higher than those already reported for other indoor environments. The plasma concentrations of both nicotine and cotinine increased significantly after the pub visit (P = 0.001 and P = 0.0007, respectively). Accordingly, the overall DNA adduct profile in induced sputum, but not in PBL, changed quantitatively and qualitatively after the pub visit. Of most significance was the formation of a distinct DNA adduct in induced sputum of three individuals consequent to ETS exposure. This adduct co-migrated with the standard (+/-)-anti-benzo[a]pyrene diol epoxide-DNA adduct, which is known to form at lung cancer mutational hotspots. We conclude that real life exposure to ETS can give rise to pro-mutagenic lesions in the lower airway, and this can be best investigated in a relevant surrogate matrix such as induced sputum.