Restoration programs need to increasingly address both the restitution of biodiversity and ecosystem services and the preparation of habitats for future climate change. One option to adapt habitats to climate change in the temperate zone is the translocation of southern populations to compensate for climate change effects - an option known as assisted migration (AM). Although AM is widely criticized for endangered species, forest managers are more confident that tree populations can be translocated with success because of previous experiences within native ranges. Here, we contend that translocations of tree populations are also subject to uncertainties, and we extract lessons for future programs of AM within species ranges from a well-documented failed case of population translocation of Pinus pinaster Ait. in Europe. The failure of these translocations originated from the unawareness of several unpredictable ecological and social events: cryptic maladaptation of the introduced populations, underestimation of climate variability differences between the source and target sites and complexity in the management schemes, postponing decisions that could have been undertaken earlier. Under the no-analog conditions that are expected with climate change, management decisions need to be made with incomplete data, implying that a certain degree of maladaptation should always be expected when restoring plant populations from local or external seed sources.