Despite the established causality between tobacco smoking and lung cancer (IARC 2004), it is difficult to establish reproducible and validated animal inhalation models for this disease. These models are needed for the evaluation of potential risk-reduced products and chemoprevention. In order to further investigate the A/J mouse model for lung tumorigenicity, male A/J mice were whole-body exposed to diluted cigarette mainstream smoke from the standard reference cigarette 2R4F for 6 hours per day, 5 days per week for 5 months followed by 4 months without exposure. The animals were exposed to fresh filtered air, to whole smoke at concentrations of 0.12 or 0.24 mg total particulate matter (TPM)/l, to gas phase depleted particle phase (PP) at 0.24 mg TPM/l, which was generated by an activated charcoal adsorber, or to the gas phase of cigarette mainstream smoke at a concentration equivalent to 0.24 mg TPM/l. Lung tumor incidence and multiplicity were determined at 5 months in the control group and the whole smoke high dose group and at 9 months in all groups. At 5 months, the tumor response was comparable in the two groups. A distinct inflammation was indicated in the whole smoke high dose group by a severe increase in neutrophils and macrophages in bronchoalveolar lavage fluid. At 9 months, both cell types returned nearly to background levels; however, lung tumor incidence and multiplicity in the whole smoke groups were dose dependently higher by a factor of up to 3-fold compared to control. The lung tumor response was similar for PP and the high concentration of whole smoke. GP at the concentration used, failed to enhance lung tumor multiplicity above control level, step-serial lung section analysis confirmed the macroscopic results. The lung tumor spectrum was the same in all groups with bronchiolo-alveolar adenoma being the most prominent lung tumor type. Further research is needed to evaluate the relevance of this model with regard to the human disease.