In summary, studies reported to date have largely failed to demonstrate human infection with animal and poultry retroviruses or an association between human diseases and these viruses. A number of studies, most of them serologic, have attempted to demonstrate human infection with these viruses. The lack of antibodies in apparently exposed groups of persons suggests an absence of infection. However, another possible explanation is that humans may be immunologically unresponsive to infection with these viruses. Although attempts to infect normal human cells in vitro with many of these viruses have not been reported, BLV and BIV appear to grow poorly or not at all. On the other hand, ALSV subgroup D infect and transform human cells in vitro. However, the production of infectious virus in vitro has been low or nonexistent. This may explain the absence of antibodies in human populations. Furthermore, many of the methods used to detect infection, either directly or indirectly, have either low sensitivity or problems with specificity. Several epidemiologic studies have tried to show a relationship between human and animal leukemia or lymphoma. In many of these studies the actual exposure to retroviruses is unknown and exposure to animals may merely represent exposure to other risk factors that are more important but were either not considered or are undefined; alternatively, a common exposure may be responsible for malignancy in humans and animals with no interspecies relationship. Based on the reported studies, these viruses appear unlikely to be responsible for any significant occurrence of human disease, particularly lymphoid malignancies. Although a definitive statement of no risk to human health is probably unwarranted, the evidence to date indicates that the risk is low and perhaps nonexistent. Thus, no specific public health recommendations regarding retrovirus-infected animals or poultry are warranted at this time.