Coronary heart disease (CHD) is the leading cause of mortality in the United States. The present cohort study was conducted to determine whether rate of FEV1 loss independently predicts CHD mortality in apparently healthy men. White male Baltimore Longitudinal Study of Aging (BLSA) participants without CHD underwent clinical evaluation at 2-yr intervals; 883 had satisfactory pulmonary and lipid studies and returned for a least one visit. Cases were BLSA subjects without CHD on entry who died a "coronary death" (death from acute myocardial infarction, sudden death, or congestive heart failure in the presence of coronary artery disease). Forced expiratory maneuvers followed American Thoracic Society guidelines. Serum cholesterol, blood pressure, cigarette smoking, and body mass index were obtained from the BLSA database. There were 79 CHD deaths and 804 survivors during an average follow-up of 17.4 yr. After adjustment for age, initial FEV1% predicted, smoking status, hypertension, and cholesterol, a time-dependent proportional hazards model showed that cardiac mortality, but not all causes of mortality, generally increased with increasing quintile of FEV1 decline for the entire cohort (relative risk [RR] 2.92-5.13) and separately for the subset of never-smokers. Thus, excess CHD mortality follows a large decline in FEV1, independent of the initial FEV1% predicted, cigarette smoking, and other common CHD risk factors.