In his introduction to Modern Practice in Tuberculosis, Gloyne (1) rightly draws attention to our comparative ignorance of the stage at which a patient with pulmonary tuberculosis becomes noninfectious. Sufficient information is not available to judge whether a smear negative for acid-fast bacilli on microscropic examination is sufficient or whether it is also necessary to have sputum negative for tubercle bacilli on culture or, in the absence of sputum, cultures of a laryngeal swab or gastric lavage. The investigation presented in this paper attempts to link and assess the degree of risk of infection with the varied sputum status of the patient. For this purpose, source cases of pulmonary tuberculosis which have come under observation during a five-year period have been divided into three groups:(I) those with sputum positive for acid-fast bacilli on microscopic examination;(^) those with sputum positive for tubercle bacilli only on culture; and (S) those with sputum negative for tubercle bacilli by both methods. When ever possible, the home contacts of these groups have been examined roentgeno graphically, and the children have been tuberculin-tested. The degree of infec tivity of the different groups of source cases has been compared by analyzing the morbidity and tuberculin state of their respective contacts.