Blue collar employees currently working in environments free from exposure to respiratory hazards were examined with chest radiography and a standard occupational history questionnaire. Workers who had worked for a total of 5 yr or more in previous jobs with possible hazardous respiratory exposures were excluded. Each radiograph was read independently by 3 NIOSH-certified "B" readers. For small opacities, the median profusion was accepted as a summary reading. The 1,422 readable films represented a population of 50.6% males, 49.4% females, 52.5% whites, 44.2% blacks, 47.0% current smokers, and 38.5% nonsmokers. The mean age was 33.8 yr, with a range from 16 to 70 yr. Small opacities of profusion greater than or equal to 1/0 were identified in only 3 (0.21%) of the radiographs--2 with small rounded opacities and 1 with small irregular opacities. Small irregular opacities of profusion category greater than or equal to 0/1 were statistically associated with age, gender, and pack-years of smoking. The results suggest that using the median of 3 independent readings should rarely result in interpretation of chest radiographs as "positive" for pneumoconiosis in active workers who have not had significant dust exposure.