1. 1) Dermal leishmaniasis (“Bay-sore”) in British Honduras is restricted to people who penetrate the forested regions. This suggests the disease to be a zoonosis, with wild animals acting as a source of infection for man. 2. 2) A variety of wild and laboratory animals were tested for susceptibility to Leishmania mexicana in attempts to indicate potential reservoir-hosts. Of the wild animals, only the cotton-rat ( Sigmodon hispidus) was readily infected. Among laboratory animals the most susceptible were hamsters, mice, gerbils and rats, in that order. It was concluded, therefore, that the reservoir-hosts of L. mexicana would best be sought among the sub-Order Myomorpha of the Order Rodentia. 3. 3) A focus of infection in the forest was pin-pointed by the isolation of L. mexicana from a batch of sand-flies trapped in a selected area. About 300 insects were triturated together in Locke's solution and inoculated into the skin of a hamster. A lesion containing LD bodies developed at the inoculation site. 4. 4) L. mexicana was isolated from three different species of forest rodents trapped in this same area: the incidence of infection was as follows: The “tree-rat”, Ototylomys phyllotis Merriam (Cricetidae, Rodentia), eight of 20 (40 per cent.) examined. The “spiny-pocket-mouse”, Heteromys desmarestianus Gray (Heteromyidae, Rodentia), six of 58 (10.34 per cent.). The “vesper-rat”, Nyctomys sumichrasti Saussure (Cricetidae, Rodentia), one of eight (12.5 per cent.). No infections could be found in a variety of other mammalian species from this and other areas of the forest. 5. 5) Infection of the reservoir-hosts appeared to be purely cutaneous and restricted to lesions on the tails. Abundant LD bodies were usually evident in stained smears from such lesions and the histology of the lesions is described. 6. 6) In all respects the strains of L. mexicana from the three rodent species appear identical with the parasite causing the human disease. Leptomonads inoculated into human volunteers produced long-lasting, cutaneous lesions containing abundant LD bodies. 7. 7) It is concluded that dermal leishmaniasis in British Honduras is a zoonosis. L. mexicana is primarily a parasite of forest rodents among which Ototylomys phyllotis, Heteromys desmarestianus and Nyctomys sumichrasti figure prominently. Other rodent species are probably also involved. Man is to be regarded as an accidental host. 8. 8) Evidence is presented to indicate that transmission among the wild animals, and to man, is by the agency of Phlebotomus species biting at ground level. 9. 9) The need is indicated for further research on the distribution and life histories of the reservoir-hosts throughout the forests. In particular a study must be made of their nesting habits to determine the possibility of a specific Phlebotomus—rodent association within the nests. 10. 10) The significance of the present findings is discussed with regard to cutaneous leishmaniasis in other parts of Central and South America. Similar investigations should be made on the epidemiology of muco-cutaneous leishmaniasis, due to L. braziliensis. There are a number of apparently distinct clinical forms of cutaneous and muco-cutaneous leishmaniasis in the New World. The possibility is considered that these might be due to infection of man with different leishmanial parasites which are normally found in different wild animal hosts.