Aim Species ranges are highly dynamic, shifting in space and time as a result of complex ecological and evolutionary processes. Disentangling the relative contribution of both processes is challenging but of primary importance for forecasting species distributions under climate change. Here, we use the spectacular range expansion (ca. 1000 km poleward shift within 10 years) of the butterfly Pieris mannii to unravel the factors underlying range dynamics, specifically the role of (i) niche evolution (changes in host‐plant preference and acceptance) and (ii) ecological processes (climate change). Location Provence‐Alpes‐Côte d’Azur, France; North Rhine‐Westphalia, Rhineland‐Palatinate and Hesse, Germany. Taxon Insect and angiosperms. Methods We employed a combination of (i) common garden experiments, based on replicated populations from the species’ historical and newly established range and host‐plant species representative for each distribution range, co‐occurrence analyses and (ii) grid‐based correlative species distribution modelling (SDM) using Maxent. Results We observed changes in oviposition preference, with females from the newly established populations showing reduced host‐plant specialization and also an overall increased fecundity. These changes in behaviour and life history may have enabled using a broader range of habitats and thus facilitated the recent range expansion. In contrast, our results indicate that the range expansion is unlikely to be directly caused by anthropogenic climate change, as the range was not constrained by climate in the first place. Main conclusions We conclude that evolution of a broader dietary niche rather than climate change is associated with the rapid range expansion, and discuss potential indirect consequences of climate change as trigger for the genetic differences found. Our study thus illustrates the importance of species interactions in shaping species distributions and range shifts, and draws attention to indirect effects of climate change. Embracing this complexity is likely the key to a better understanding of range dynamics.