Although smoking is associated with cardiovascular disease (CVD), many individuals remain healthy after many years of smoking. The population based cohort 'Men born in 1914' was used to investigate whether the occurrence of non-invasively detected atherosclerosis among smokers is associated with lung function [(i.e. height-adjusted forced expiratory volume during 1 s (FEV1.0) and vital capacity (VC)]. Two hundred and seven smokers without history of CVD were examined with spirometry and calf plethysmography at 55 years, and with spirometry, ankle-arm blood pressure recordings and ultrasound examinations of the carotid arteries at 68 years. Eighty-three men had atherosclerosis defined as carotid stenosis >30% or ankle-arm index <0.9. FEV1.0 and VC were both at 55 years (longitudinally) and at 68 years (cross-sectionally) lower among men with atherosclerosis at 68 years (55 years: FEV1.0, 3.2+/-0.6 vs. 3.4+/-0.5 l; P=0.02; VC, 4.2+/-0.5 vs. 4.4+/-0.5 l; P=0.02; 68 years: FEV1.0, 2.6+/-0.6 vs. 2.9+/-0.7 l; P=0.004; VC, 3.8+/-0.6 vs. 4.0+/-0.6; P=0.009, for men with and without atherosclerosis). The longitudinal and cross-sectional associations between FEV1.0, VC and atherosclerosis remained significant after adjustments for several potential confounders (tobacco consumption at 55 and 68 years, hypertension, diabetes, alcohol consumption at 68 years, and pulse wave amplitude as a measure of degree of atherosclerosis at 55 years). We conclude that the risk of developing atherosclerosis is associated with the degree of ventilatory capacity. The results suggest that in smokers, reduced lung function is a marker of susceptibility for atherosclerosis.