New composite ('quadrant') standard films for classifying radiographs of pneumoconioses

120 chest radiographs of dust exposed persons were classified twice by each of 39 physicians, and three times by 37 of them. All film readers used both the 22 standard films accompanying the current International Labour Office's (1980) Classification of Radiographs of Pneumoconioses, and a modified set that reproduced selected parts from some of them as quadrants of full-size films. Variability within and between readers in small opacity profusion classifications was high, but similar when using the two sets of standards. Some readers recognised irregularly shaped small opacities more frequently, in low profusion categories, when using the quadrant standards. Small rounded opacities were classified more frequently higher on the scale when using the current standards. The effects found were small although the test films included a high proportion with some small opacities. It is concluded that use of the new quadrant standards is unlikely to introduce any important bias into results from occupational health surveys.