Urban impacts on climate are generally discussed at the local scale (urban heat islands), with urban areas ignored in large-scale climate assessments and for future climate projections since cities historically covered a small fraction of the Earth’s surface. Combining millions of satellite-derived images of land surface temperature with dynamic urban area estimates, we demonstrate that the urban influence on continental- to regional-scale warming has become stronger over time, especially for rapidly urbanizing regions and countries in Asia. However, the main cause of global warming is still not urbanization, contributing to only around 2% of the land warming during the study period. We also estimate these large-scale urban warming signals under all shared socioeconomic pathways used to project global and regional climate change. Based on these results, we argue that, in line with other forms of land use/land cover change, urbanization, or rather urban evolution, which incorporates both changes in urban extent and properties of urban land over time, should be explicitly included in future climate change assessments across scales. These patterns, also seen for air temperature, reframe our understanding of urbanization in the climate system from only a local-scale phenomenon to one with non-negligible regional- and even continental-scale impacts. Finally, we examine the impacts of choice of dataset, particularly the urban land cover products, on this and other use cases, showing that, while the signals themselves are usually consistent, the magnitude of the signal can strongly depend on the urban land product. This sensitivity analysis demonstrates the importance of choosing fit-for-purpose datasets for examining specific aspects of historical, present, and future urbanization with implications for sustainable development, resource allocation, and quantification of climate impacts.