The relation of body mass index, based on self-reported weight 5 years prior to diagnosis, to lung cancer was investigated in 3,607 lung cancer cases and 9,681 controls interviewed in a hospital-based case-control study in eight US cities between 1981 and 1990. Separate analyses were carried out by smoking status and by sex. After adjustment for covariates, odds ratios for lung cancer by levels of body mass index, taking greater than or equal to 28 as the referent, showed an increasing linear trend with decreasing body mass index for current smokers and ex-smokers of both sexes and for female never smokers. These results are consistent with findings of a number of prospective studies. Further studies are needed to determine whether the association of low body mass and lung cancer is due to factors associated with leanness or to a biologic effect of leanness itself.