Because pathologists differ widely in their definition and histologic classification of many types of cancer, progress has been impeded in determining the frequency, nature, and biological behavior of tumors, in elucidating their causes, and in evaluating treatment and prognosis. Histologic definitions and systems of classification that are accepted and utilized internationally are essential to advancement of our scientific information in many aspects of cancer. I n order to make meaningful comparisons of data there must be reasonable uniformity in histologic definitions and classifications. Although the disparity in use of different and varying classifications for some tumors is generally recognized, recently more quantitative information has been published for lung cancer. Sobin" reviewed 57 conveniently available English language papers on lung tumors between 1955 and 1970, and found 13 different classifications used. The 34 papers concerned with treatment and survival used 12 different classifications, and the 23 papers concerned with clinical diagnosis, cytology, epidemiology, genetics, occupation, and pathology used 7. Few papers defined the criteria