The presence of bovine tuberculosis ( Mycobacterium bovis) in the Kruger National Park (KNP) was determined for the first time in 1990. It was diagnosed in an African buffalo ( Syncerus caffer) bull, which was found recumbent and in an emaciated and moribund state near the south-western boundary fence. This prompted an investigation into the bovine tuberculosis (BTB) status of the KNP, with emphasis on its epidemiological determinants and risk factors. This report documents the findings of surveys that were conducted from 1990 to 1996. It was found that BTB had entered the KNP ecosystem relatively recently (±1960), and has found favourable circumstances for survival and propagation in a fully susceptible and immunologically naive buffalo population. Indications are that it entered the KNP from across the southern river boundary, where the presence of infected domestic cattle herds had been documented. From there the infection spread through the southern buffalo population and is currently spreading in a northward direction. It was estimated that this northward spread took place at a rate of about 6 km per year; the prospect being that, if this rate of spread is maintained, the entire KNP may be affected in less than 30 years from now. Spillover from buffalo had already occurred in species such as chacma baboon ( Papio ursinus), lion ( Panthera leo), cheetah ( Acinonyx jubatus), kudu ( Tragelaphus strepsiceros) and leopard ( Panthera pardus). Although there is no indication yet that these species act as maintenance hosts, the possibility is raised that these, or an as yet overlooked species, might assume such a role in future. In the KNP, BTB manifests itself as a chronic and predominantly subclinical disease in buffalo. It may take years for clinical signs to develop, and then only at a terminal stage, when emaciation is a constant feature. It is suspected that the time from infection to death is variable and dependent on the animal's immune response, which can be weakened by such factors as stress, old age or droughts. It was found that, in the interim, buffalo have a normal reproductive life. On necropsy, buffalo show almost exclusively lung and upper respiratory tract involvement, pointing to an aerogenous mode of transmission. Histologically, little sign of encapsulation of lesions was detected, which suggests that they are exceptionally susceptible to BTB and that most lesions are open and infectious and progressive, leading ultimately to death of the individual. Evidence also indicates that BTB is progressive with the herd context (92% being the highest prevalence rate thus far determined in a buffalo herd) as well as progressive within the KNP buffalo population (the implication being that virtually all buffalo herds in the KNP will eventually be infected). Preliminary data suggest a positive correlation between disease prevalence and mortality, with potential mortality reaching up to 10% in buffalo herds having BTB prevalence rates of 50% and higher. Only the future will tell what the effect of the disease on the population dynamics of buffalo will be.