There has been a tremendous change during the past few years in the manner of treatment of tuberculosis. Stillthe presence of a cavity in pulmonary tuberculosis is of extreme importance for the treatment and furtherprognosis of the disease. Any method of examination which helps to diagnose cavities more accurately is ofconsiderable value. Body section radiography or tomography, orginally discovered by Bocage in 1921, is such amethod, but it should by no means be considered the "last word" in the detection of cavities. Still there remainscertain problems that must be solved by further studies. Our study is to check the accuracy of the tomographicinterpretations by correlating them with the actual lung findings in 63 cases which received either pneumonectomyor lobectomy at the Dept. of Thoracic Surgery, Severance Hospital, Yonsei University, Seoul, Korea. Results may bebriefly summarized as follows; 1. 30% of cavities clearly visualized on P-A chest films were confirmed by theresected specimens and among the suspected cases, 39.5% were confirmed to have cavities, a total of 69.5% thusbeing visible. 2. Of the 41 cases predicted to have cavities, 39 were confirmed by the resected specimen, or avisibility of 85%. Of the 7 suspected cases, 5 were confirmed, the possibility of diagnosis among suspected casesbeing 10.6%, a total of 95.6% thus being visible. 3. The limitation of tomography was discussed and some of thepitfalls in diagnosis were incicated. Particular emphasis is placed on emphysematous belbs and bronchiectaticdilatations which may be extremely difficult to distinguish from small tuberculous cavities, especially those inthe extreme apical portion of lung.