1. 1. The accuracy of the properly performed tuberculin skin test beginning with 1–1000 O.T. and followed by 1–100 O.T.), if negative with both dilutions, is so great in ruling out tuberculosis that it becomes the keystone in the diagnosis of the disease, since 59 per cent of patients even with roentgenographic changes sufficient (in miniature films) to arouse suspicion of tuberculosis have negative Mantoux tests. 2. 2. A positive Mantoux test in a child less than six years of age indicates active primary disease. A positive Mantoux test in a patient over six years of age indicates a prior primary infection, and if disease is present anywhere in the body, tuberculosis must be ruled out. 3. 3. In order to recover tubercle bacilli from lesions, repeated specimens cultured on two different media are required because the tubercle bacilli may not be present in sputum or gastric cultures every day and may grow better in one medium than another even though both will support tubercle bacilli in the positive control. 4. 4. Lowenstein's and Petragnani's media are complementary and equally effective in outpatient cultures. Petragnani's medium gives a higher percentage of positive results on inpatient cultures and a lower percentage of contaminations. 5. 5. A single positive smear or culture is sufficient reason to hospitalize a patient and to begin treatment. 6. 6. Many factors influence the recovery of tubercule bacilli: (a) total number of tubercle bacilli in the specimen; (b) the digestant material kills 80 per cent of the viable tubercle bacilli; (c) pH above or below 7 may rapidly kill the tubercle bacilli; (d) temperature changes of specimens above or below 30 °c. decrease the number of viable tubercle bacilli; (e) by-products of vitamin C in urine are bactericidal; and (f) antibiotics inhibit the growth of tubercle bacilli on culture media. 7. 7. Contamination is a major problem in culture work, the highest contamination occurring in outpatient specimens, drainage from sinus tracts, and aspirated fluids. The lowest contamination rate occurs with inpatient sputums. 8. 8. Using three specimens cultured on two different media, we have recovered tubercle bacilli from 22.5 per cent of primary cases, 94 per cent of patients with minimal tuberculosis, 95 per cent of patients with moderately advanced pulmonary tuberculosis, and 95 per cent of patients with far-advanced tuberculosis. 9. 9. The erythrocyte sedimentation rate is of no value in the diagnosis of active tuberculosis since we found it to be normal in over one-fourth of patients with positive cultures. 10. 10. A pathologic diagnosis of tuberculosis was not returned in thirty of 189 specimens sent even though one or more positive cultures had been obtained from each patient. There was disagreement between culture and pathologic diagnosis in 15.8 per cent of these cases. 11. 11. Since the advent of combined chemotherapy all our tuberculosis deaths occurred in patients with advanced disease, either pulmonary or extrapulmonary, which was present on admission; there were no deaths from primary or minimal pulmonary disease. 12. 12. When one realizes the difficulties in recovering tubercle bacilli and the ways in which these procedures can go amiss, the value of short-term work-up with three gastric specimens that can be cultured immediately by qualified technicians using culture media that grow tubercle bacilli, becomes of increasing importance. If the bacilli are dead on arrival, or viable tubercle bacilli are planted on culture media that will not support them, negative results mean nothing. 13. 13. Without the three-day period of hospitalization during which three gastric specimens are cultured, the diagnosis would have been missed on one-third of the total group; early disease, minimal and moderately advanced pulmonary and primary disease would have been missed most commonly. 14. 14. Tissue examination is of great value in the diagnosis of tuberculosis in extrapulmonary lesions since one-third of this group were diagnosed on the basis of pathology reports; it is of less value in diagnosing moderately and faradvanced disease and of no value in minimal and primary disease since these patients do not require resection nor do they die of the disease. 15. 15. Active tuberculosis is a clinical and not a pathologic diagnosis and it is best substantiated by the recovery of tubercle bacilli from the lesion.