Current changes in island vegetation are caused among others by the introduction and spread of non-native plant species. While current invasion statuses are well-known and insular biodiversity is being monitored, a paleoecological perspective on the timing, trajectories, and magnitude of vegetational change on islands caused by non-native plants is largely missing. By matching long-term low-resolution pollen data with short-term high-resolution plant data, we quantify the changes caused by non-native plants on islands globally during the last 5000 years. In general, non-native plants have been present in many systems but started to increase massively only during the last 1000 years, without any signs of slowing down. Thus, although humans have altered the vegetation on islands for centuries and millennia, the recent increase in non-native plants is unprecedented. We show that palynological data can be used to obtain a historic perspective on the development of non-native vegetation on islands.