We studied dietary risk factors for lung cancer among never-smokers in a population-based case-control study in Stockholm, 1989-1995. Study subjects were older than 30 years of age and had never smoked regularly. A total of 124 cases (35 men, 89 women) and 235 controls (72 men, 163 women) participated. Exposure information was obtained at interview with study subjects. The never-smoking status was validated by interviews with next-of-kin. A protective effect was suggested for vegetables, mediated primarily by carrots (relative risk [RR], 0.7; 95% confidence interval [CI], 0.4-1.3, and 0.6, 0.3-1.1 for intermediate and high consumption of carrots, respectively). Non-citrus fruits appeared to lower the risk as well, with RR 0.6, 95% CI 0.3-1.3 and 0.5, 0.3-1.0 for intermediate and high consumption, respectively. A protective effect with dose-response was also seen for intake of beta-carotene and total carotenoids. Increased risks were seen for cultured milk products in both genders (RR 2.0, 95% CI 1.1-3.9 for intermediate and 1.6, 0.9-2.9 for high consumption), but for milk only among male high consumers. Our results support evidence linking a diet rich in vegetables and non-citrus fruit with decreased lung cancer risk and suggests that among vegetables, carrot consumption is the most important component or marker for this effect in Sweden. The results regarding milk products could be consistent with dietary fat as a risk factor for lung cancer, although a more comprehensive assessment of fat intake is necessary to explore this relation.