Climate change is challenging the predictive ability of ecologists, evolutionary biologists and scientists as a whole. We are urged to anticipate the impacts in order to attempt mitigation, but reliable forecasts of the future dynamics of biodiversity are elusive. In contrast, retrospective studies can provide key insights based on a better knowledge of species responses to past climate changes. To date, this retrospective knowledge has mostly been based on the fossil record, genetic data (phylogeography) and hindcasting of species distribution models. Despite the invaluable contributions of these approaches, all of them suffer from particular biases and assumptions that limit the reach of their inferences. A rigorous way to overcome this problem is to integrate multiple data sources into a single statistical framework that explicitly accounts for the biases and intrinsic limitations of different disciplines. Upon this rationale, we have developed a dynamic, process-based model of species range dynamics that, by means of Bayesian data assimilation, exploits information from multiple sources. This quantitative framework is enabling us to better understand how European trees responded to past climate changes in terms of migration, regional extinction or persistence in previously unknown refugia. Hence, integrative reconstructions of the past represent a promising way to promote ecological understanding and improve predictions of future impacts of climate change on Earth’s biodiversity.