OBJECTIVE The relation between pulmonary function and low attenuation areas (LAAs) on high-resolution computed tomography (HRCT) is not clear in subclinical pulmonary emphysema. Accordingly we examined pulmonary function and HRCT in asymptomatic community-based volunteers. METHODS The existence of LAAs on CT was evaluated independently by three respiratory physicians who were blind to the pulmonary function test data and smoking histories of the subjects. The LAA grade was assessed by a visual scoring method from 0 to 5, and the individual LAA score and the values of pulmonary function tests were compared. METHODS 57 subjects aged from 32 to 82 years [mean 62 +/- (SD) 11] years were enrolled in the study. RESULTS LAAs were found in 20 of the 49 subjects who were current or ex-smokers but in none of the 8 who never smoked. LAAs observed were mostly less than 25% of the total areas of the lungs. Although all mean values were within normal limits, the forced expiratory volume in 1 s (%pred), maximal expiratory flow at 50% of forced vital capacity (%pred), diffusing capacity of the lungs for carbon monoxide (DLCO; %pred) and diffusing capacity for carbon monoxide per unit alveolar volume (DLCO/VA; %pred) were all significantly less in those with LAAs than in those without LAAs. There was also a significant tendency for DLCO %pred and DLCO/VA %pred to decrease as the LAA scores increased. CONCLUSIONS Those who have LAAs on CT have a small but significant deterioration in pulmonary function tests even if the LAAs observed are minimal or mild in degree. Longitudinal studies will be necessary to determine whether LAAs on CT identify the subjects who will develop airflow obstruction over time.