We examined the lungs of 48 patients who died in the National Institutes of Health clinical trial of Intermittent Positive-Pressure Breathing. Bronchial lesions were not related either to clinical features or to tests of expiratory flow. In particular, bronchial muscle was not related to the response to bronchodilators or to the variability of flow rates. The degree of emphysema was well related to loss of body weight and was also positively related to right ventricular hypertrophy. It was also the major morphologic correlate of abnormal tests of expiratory flow, the slope of phase III of the single-breath nitrogen test, and increased residual volume. Emphysema was negatively related to bronchodilator response and variability of expiratory flow. The proportion of bronchioles less than 400 microns in diameter and a measurement of irregular bronchiolar shape were related to dyspnea and increased PaCO2 and were also related to abnormal tests of expiratory flow, phase III of the single-breath nitrogen test, and residual volume. The presence of increased muscle and fibrosis in the bronchioles appeared beneficial, being associated with a higher PaO2, a lower PaCO2, less edema, less right ventricular hypertrophy, better flow rates, and lower residual volumes. This may reflect the fact that airways with increased muscle and fibrosis are less markedly narrowed and less distorted. Goblet cell metaplasia appeared to be an important variable, independent of others, and was related to airway resistance. We divided patients into 3 groups according to severity of emphysema, and then into 2 subgroups of lower and higher flow rates in each category.(ABSTRACT TRUNCATED AT 250 WORDS)