In order to learn whether the neonatally thymectomized Lewis rat (NTLR) infected with Mycobacterium leprae could serve as a model for chemotherapeutic studies in a situation resembling that found in human lepromatous leprosy, NTLR inoculated with M. leprae either locally or intravenously 9 to 16 months earlier were treated for from 1.5 to 8.5 months with dapsone (4,4'-diaminodiphenylsulfone, DDS) incorporated in the rat chow in the concentration providing the minimal inhibitory concentration of the drug for M. leprae and in the 100-fold larger concentration. NTLR were killed at intervals; the M. leprae were counted and passed to mice. Treatment with the smaller dosage of dapsone neither killed M. leprae nor reduced the number of organisms in the bacterial populations, whereas treatment with the larger dosage both killed M. leprae and reduced their numbers. The rate at which the organisms were killed (i.e., rendered noninfective for mice) was much the same as that in patients treated with dapsone in comparable dosage. The dead organisms were removed from the rat tissues at a faster rate than encountered in patients. The NTLR may indeed be suitable for chemotherapeutic studies relevant to man. In addition, the more rapid disappearance of dead M. leprae from the rat tissues may facilitate the study of treatment regimens designed to eradicate persisting viable organisms.