Follow-up of the prevalence study of respiratory symptoms and chronic airway obstruction was performed after a 6- to 7-year interval. One hundred fifteen of 117 subjects (98%) originally labeled "abnormal" (chronic bronchitis or asthma by history, or ratio of 1-sec forced expiratory volume to forced vital capacity less than 60 per cent), and 111 of a random sample of 116 subjects (96%) originally labeled "normal" were traced. Nineteen patients were dead, 14 "abnormal" subjects, and 5 "normal" subjects (P less than 0.01). Eighty-eight of the originally labeled "abnormal" subjects and 91 of the random "normal" sample could be retested by spirometry. Subjects with lower initial ratios of 1-sec forced expiratory volume to forced vital capacity tended to lose more 1-sec forced expiratory volume and forced vital capacity. Early identification of respiratory symptoms and spirometric abnormalities may identify persons at increased risk of death who could benefit from appropriate therapy.