A proven case of cryptococcosis was seen in the Department of Dermatology, University of Michigan Hospital, and attempts were made to trace the possible source of infection. At the onset of his illness, the patient was working as a janitor in the Foster School of Ypsilanti, Michigan. This is a rambling, one-story schoolhouse built during the Second World War. In view of the recent reports of the isolation of Cryptococcus neoformans in nature (1, 4, 5), samples of soil and floor dust from the schoolhouse were obtained: Sample 1 and Sample 2 were soils from under drains, Sample 3 was from the sweepings gathered in the bag of a vacuum cleaner used to clean the floors, and Sample 4 was from dirt accumulated on a motor-driven rotary floor brush. Each sample was suspended in physiological saline solution in a 100 ml graduate with vigorous agitation and left to settle at room temperature. After 3 hours, portions of the top layers of the suspensions were inoculated on duplicate Sabouraud's-dextrose-agar slants, both with and without penicillin and streptomycin. One set of the cultures was left at room temperature and the duplicate set, at incubator temperature. These were examined periodically for growth and discarded after six weeks. For animal inoculation, one ml of the supernatant layer of each suspension was injected intraperitoneally into mice. Each suspension was used to inoculate five mice. Pieces of the livers, spleens and lungs of animals that died were planted on Sabouraud's dextrose agar slants. The remaining animals were sacrified after three weeks and the organs similarly planted. The cultures directly planted with the samples did not show any growth of pathogenic fungi, and most eventually became overgrown with contaminants. However, Trichophyton mentagrophytes was isolated from the livers or spleens of two animals inoculated with portions of Sample 3, composed of vacuum cleaner sweepings. The macroscopic and microscopic features of these isolates were characteristic (FIGS. 1,