The value of mucus hypersecretion as a predictor of mortality and hospitalization was studied in a random population sample of 876 men, aged 46-69 years. The cohort was examined in 1974 with the British Medical Research Council questionnaire and lung function tests. A total of 219 men had died between 1974 and 1985. Twenty-seven men died from lung cancer and 14 died from other respiratory diseases. Mucus hypersection was not found to be significantly related to overall mortality after controlling for age, smoking and FEV1. Similarly, mucus hypersection was not a predictor of lung cancer mortality after controlling for age and smoking habits. The predictive value concerning death due to respiratory disease could not be examined because of the limited number of deaths in the cohort from these diseases. Mucus hypersecretion was not significantly related to hospitalization in general. Mucus hypersecretion had a significant predictive value concerning hospitalization due to respiratory disease in general, but the value was insignificant after controlling for FEV1. In contrast to this, mucus hypersecretion was a significant predictor of hospitalization due to COPD, even after controlling for FEV1. We conclude that the predictive value of mucus hypersecretion concerning mortality is of no value. Concerning morbidity, our results show that, although secondary to airflow obstruction, mucus hypersecretion must be viewed as an indicator of severity of COPD.