Modelling of emerging vector borne diseases serves as an important complement to clinical studies of modern zoonoses. This article presents an archaeo‐historic epidemiological modelling study of Rift Valley fever (RVF), using data‐driven neural network technology. RVF affects both human and animal populations, can rapidly decimate herds causing catastrophic economic hardship, and is identified as a Category A biodefense pathogen by the US Center for Disease Control. Despite recent origins circa the early 1900s, little is known about the circumstances of its inception nor the relationships between factors that affect transmission. This evidence could be vital as the disease continues to expand from its epicentre in Kenya to other parts of Africa and the Arabian Peninsula. RVF is a relevant case for archaeological/palaeopathological investigations of disease as it intersects between numerous human, animal, spatial, temporal, and sociopolitical dimensions. By integrating landscape archaeology, historical evidence, and climatic data, with evidence of human behaviour gathered through ethnoarchaeological study, this article presents an applied framework for human–animal palaeopathology. This framework aligns with the One Health approach that observes disease to be intrinsically tied to ecological and societal factors. We provide a useable alternative way of thinking about disease modelling in the present and the past, ultimately seeking to support efforts to accurately predict future impacts. Tapping into longitudinal evidence from the last 50–300 years offers a powerful way to respond to the threat zoonoses will pose to human populations around the world as the climate warms.