We have evaluated a series of chemopreventive agents in an animal model of tobacco smoke-induced lung cancer. During 5 months, strain A mice were exposed for 6 h/d, 5 d/wk to a mixture of 11% tobacco mainstream smoke and 89% tobacco sidestream smoke, followed by a 4-month recovery period in air. The protocol has been found consistently to increase the multiplicity and incidence of pulmonary adenomas and adenocarcinomas. Feeding diets containing a mixture of dexamethasone (0.05%) and myoinositol (1%) prevented the development of lung tumors, even if treatment was began only after cessation of smoke exposure. However, such putative chemopreventive agents as phenethylisothiocyanate, benzyl isothiocyanate, d-limonene, acetylsalicylic acid, 1,4-phenylenebis (methylene) selenoisocyanate, and green tea had no protective effect.1The antioxidants N-acetylcysteine and β-carotene also failed to modulate lung tumor development.2The model would thus have predicted the negative outcome of two major clinical trials. We also found that filtered tobacco smoke, devoid of most polycyclic aromatic hydrocarbons and nitrosamines, produces lung tumors as effectively as does full smoke, casting some doubt on the general assumption that polycyclic aromatic hydrocarbons and nitrosamines are the most important tobacco smoke carcinogens.3–4 We concluded that chemopreventive agents should be evaluated for their possible effect against the full complex mixture of tobacco smoke and not just against selected single chemicals assumed to be major tobacco carcinogens.