Since October, 1956, a quantitative study of air-borne infection has been in progress at the Veterans Administration Hospital in Baltimore. The frequency of occurrence of tuberculous infections in guinea pigs eJ.: posed to air exhausted from a six-bed ward of patients with active pulmonary tuberculosis has been determined. The experimental unit is an isolated ward of six beds, with a controlled ventilating system; a constant quantity of air is continuously removed from the ward and, just before being exhausted from the hospital, is passed through a large animal exposure chamber. Riley and associates have reported the details of the method (1-3). It was not anticipated that the detailed bacteriology involved would be of special interest. However, in the course of the study it became evident that a very few of the many participating patients were causing these infections and that the majority of the animal infections were caused by drug-resistant bacilli. This paper is a retrospective analysis of these findings.