Fourteen years' experience with pneumonectomy in the treatment of pulmonary tuberculosis is reviewed. Sixty-three operations, constituting approximately 0.5 per cent of all resections for tuberculosis, are reviewed. Forty-nine (78 per cent) of the patients are living and well, and 14 (22 per cent) are dead. The mast common cause of mortality was cardiorespiratory failure which is believed to reflect not only the amount of tisue removed, but also the extent of the original disease. Bacteriologic, physiologic and clinical data suggest that pneumonectomy, with rare exceptions, should be reserved for the patient with destroyed lung who harbors organisms resistant to SM, PAS and INH.