Climate variability has been causally linked to the transformation of society in pre--industrial southeast Africa. A growing critique, however, challenges the simplicity of ideas that identify climate as an agent of past societal change; arguing instead that the value of historical climate--society research lies in understanding human vulnerability and resilience, as well as how past societies framed, responded and adapted to climatic phenomena. We work across this divide to present the first critical analysis of climate--society relationships in southeast Africa over the last millennium. To achieve this, we review the now considerable body of scholarship on the role of climate in regional societal transformation, and bring forward new perspectives on climate--society interactions across three areas and periods using the theoretical frameworks of vulnerability and resilience. We find that recent advances in palaeoclimatology and archaeology give weight to the suggestion that responses to climate variability played an important part in early state formation in the Limpopo valley (1000--1300), though evidence remains insufficient to clarify similar debates concerning Great Zimbabwe (1300--1450/1520). Written and oral evidence from the Zambezi--Save (1500--1830) and KwaZulu--Natal areas (1760--1828) nevertheless reveals a plurality of past responses to climate variability. These were underpinned by the organization of food systems, the role of climate--related ritual and political power, social networks, and livelihood assets and capabilities, as well as the nature of climate variability itself. To conclude, we identify new lines of research on climate, history and society, and discuss how these can more directly inform contemporary African climate adaptation challenges.2