A population group, comprising 653 men and 823 women aged 40-64 years, was surveyed in 1961 to determine the prevalence of chronic respiratory disease. Questionnaire, simple spirometry and thoracic radiography were used. The 10-year mortality of the population has been related to the initial findings. After standardization for age and smoking, men in each of the disease or symptom groups had a higher risk of dying during the 10-year period than men without lung disease; for the total group of men with chronic non-specific lung disease (CNSLD) the standardized risk ratio was 1.3. After standardization for age and presence of CNSLD, men in all smoking groups had a higher risk of dying than male non-smokers. For the total before the age of 15 therisk was 4.1-fold. When men with and without CNSLD were analysed separtely in different smoking groups, the excess mortality of men with CNSLD varied from 10 to 110% (risk ratios from 1.1 to 2.1), being highest among the heaviest smokers. In men without lung disease, the smokers had a 1.8-fold mortality as compared with non-smokers. In women, the CNSLD did not increase mortality, but in female smokers the risk of dying was roughly twice that of female non-smokers although the difference was not significant. It appears from the findings that, as far as the prognosis for life is concerned in the male population surveyed, smoking is more important than CNSLD. The effect of CNSLD seemed to be stronger in men who smoked at least 25 g of tobacco daily, than in men who smoked less. The ex-smokers resembled non-smokers more than smokers in mortality experience which suggests that the deleterious effects of smoking may disappear after smoking has been discontinued.