Risk factors for lung cancer among women who had never smoked were assessed in a case-control study of 161 newly diagnosed histologically confirmed cases and 483 population controls between 1994 and 1997 in eight Canadian provinces. Measurement included socio-economic status, smoking habits, alcohol use, diet, residential and occupational histories and exposure to environmental tobacco smoke (ETS). Dose-response associations were observed for consumption of tea, adjusted odds ratios (ORs) 0.6 (95% confidence interval (CI) = 0.3-0.9) for 1-7 cups per week and 0.4 (95% CI = 0.2-0.7) for > or = 8 cups per week (P = 0.0008), and smoked meat, adjusted ORs 1.3 (95% CI = 0.8-2.3) for 0.5 slice per week and 2.1 (95% CI = 1.1-4.0) for >0.5 slice per week (P = 0.02). Regular use of shortening in cooking was also related to lung cancer. Increased ORs with borderline significance were found for total consumption of meat, eggs or French fries and fried potatoes. Passive exposure to ETS at home (or at work) may be associated with lung cancer risk among never-smoker women; the adjusted ORs were 0.7 (95% CI = 0.2-2.3), 1.2 (95% CI = 0.4-3.2), 1.5 (95% CI = 0.5-4.0) for 1-16, 17-30, and 31 or more years of combined residential and/or occupational ETS exposure, respectively, with a similar pattern for smoker-years of ETS exposure.