The perception of abnormal shadows is fundamental to roentgenological diagnosis. A lesion must be greater than certain limits in size tobe roentgenologically visible. On the other hand, considerably larger lesions may frequently escape from being detected when they are located in so-called blind areas of the lung fields. There seems to be relatively a high intraobsever and interobserver error in detecting and describing abnormal pulmonary shadows. On this occasion, a brief comment has been made on the variability and limitation in perception of pulmonary lesions as well as on the advantages and shortcomings of dual interpretation. A few measures which may be of alue in redcing the undesired unreliability of chest X-ray interpretation has also been shortly discussed.