Abstract A growing scholarship is focusing on the cost of social resilience to climate change in the past. Among different resilience strategies, migration could be effective for nomadic societies despite the potential consequences of conflicts and epidemics. Thus, this study utilizes historical records to statistically investigate the linkages among nomadic migrations, nomad–farmer conflicts, and epidemics under climate change and population pressure in imperial China (200 BCE–1840 CE) on the national and provincial scales. The current study will first attempt to empirically identify and analyze the cost of resilience to climate change mainly in the direction from nomadic societies to agrarian societies in historical China. In particular, we show the cost of nomadic migration passed in a chain mechanism as ‘climate change → nomadic migration → nomad–farmer conflicts → epidemics.’ Nomad–farmer conflicts were one direct effect of nomadic migration, while epidemics were an indirect one. Spatially, more provinces were affected under the direct effect than under the indirect effect. Furthermore, the first level of chain ‘nomadic migration → nomad–farmer conflicts’ covers more provinces than the second level ‘nomad–farmer conflicts → epidemics’. These empirical results remind us to identify and avoid the cost of resilience as early as possible before the cost may transmit further in a chain manner. However, the provinces outside the concentrated nomad–farmer conflicts did not demonstrate significant linkages between conflicts and epidemics, which highlights the importance of peaceful cross-civilizational and inter-societal interactions against common challenges of climate change. This study with a cross-scale perspective in geography provides a theoretical implication to improve the current understanding on climate justice and have a practical value to avoid or minimize the cost of resilience.