The impact of humans on climate and landscape is obvious today when anthropogenic climate change impacts are being felt around the world and more than 10% of the land surface is intensively managed. However, transformations of natural ecosystems by humans began with the shift from hunting and gathering to cultivation during the Mesolithic and Neolithic periods. There have also been substantial changes in climate during the Holocene, giving rise to natural changes in vegetation cover and fire regimes. It is important to quantify the impact of natural and anthropogenic influences on landscapes at a regional scale, and understand how climate-induced changes affected the resources available to people and hence the expansion of human populations, to provide a firm foundation for interpreting human development in the past and for predicting human-environment interactions in the future. Several recent developments make it possible to address these issues more systematically. New statistical tools have been developed that provide independent and robust quantitative reconstructions of past changes in climate, vegetation and fire regimes through the Holocene. These reconstructions can be compared with evidence for population growth and/or agricultural practices at a regional scale to isolate human and natural impacts on ecosystems and disturbance regimes. Process-based modelling tools, including eco-evolutionary optimality models of gross primary production and of specific crop types, can be used to disentangle the roles of climate, climate variability, changes in atmospheric CO 2 levels, and agricultural practices on agricultural yields through time. Transient climate model simulations can be used to examine the response of regional climates to changes in external forcing through time and also the interaction between long-term trends in climate, changes in climate modes and unforced climate variability. Together, these tools can be used to foster a better understanding of the climate-human nexus during the Holocene.