A retrospective analysis has recently been made of 1504 patients with active advanced pulmonary tuberculosis admitted to Trudeau Sanatorium 1930-1939 and followed to death or 1953-1954 in over 90 per cent.’ These patients were rather select, not only from a socio-economic viewpoint, but those with major nontuberculous complications-advanced emphysema, cardiac decompensation, diabetes mellitus, malignancy, bronchial asthma and silicosis-and major tuberculous complications-bone and joint, genitourinary, peritonitis or miliary-were excluded. During an analysis of 25 years experience with a routine admission barium motor meal in search of intestinal tuberculosis, a possible relationship between the type and distribution of pulmonary tuberculous lesions and the presence of intestinal tuberculosis was suggested.2 As a consequence, for this purpose, during the study of advanced pulmonary tuberculosis, the cases were classified in two additional ways: 1. Presence or absence on admission of bilateral symmetrical nodular distribution of the abnormal roentgenographic pulmonary shadows.3 To be so classified there must have been diffuse finely nodular shadows, most intense at the apices, gradually disappearing toward the bases and approximately equal in distribution and in intensity in the two lungs. While this picture approximates that of miliary tuberculosis, the two conditions are readily distinguished on both clinical and roentgenographic grounds. An example of the type of case being discussed is shown in Figure 1. 2. Presence or absence on admission of roentgenographic evidence of intestinal tuberculosis. For 25 years, between 1924 and 1949, about 98 per cent of patients admitted to Trudeau Sanatorium had a routine barium motor meal regardless of the pulmonary findings. Evidence of multiple areas of obstruction, rapid passage of barium through the terminal ileum and cecum, and deformity, spasm and frank ulceration of the terminal ileum and cecum were sought. There were 1504 patients on whom the 20 year morbidity and mortality statistics were available; “five year results” were available on 1429. The five year result was considered “good” if the disease had become clinically controlled with no evidence of cavity and with negative sputum bacteriology within three years and if relapse had not occurred within the balance of *From the Trudeau-Saranac Institute, Trudeau, New York. **Thjs study was made possible by a grant-in-aid from the Committee on Medical Research of the American Trudeau Society, Medical Section of the National Tuberculosis Association. f Present address: The Colorado Foundation for Research in Tuberculosis, University of Colorado Medical Center, Denver, Colorado. 669