The influence of non-native pollen, both long-distance transported and from introduced taxa, on reconstruction of past vegetation is not often well quantified in palynological investigations. We examined both fossil and modern samples from the Galápagos Islands, a remote archipelago lying 1000 km from the nearest continent. These islands are particularly well-suited for such an assessment, as (1) the native flora is limited and well-known, enabling increased taxonomic resolution within the palynological record, and (2) human impact in the Galápagos started after discovery by Europeans in 1535, allowing clear distinctions to be made between native and introduced taxa. Pollen samples were collected from five profiles in the Galápagos and grouped in (a) a pre-human-impact period, (b) an early human-impact period after c. 1535, and (c) a late human-impact period after c. 1973 when the introduced Cinchona pubescens tree started to expand. Introduced taxa accounted for approximately 10% of total pollen (excluding Cyperaceae) throughout the human-impact periods and long-distance transported pollen for approximately 5%. Twenty pollen taxa of introduced plants were found. Cinchona, which grows abundantly near the study sites, accounted for most of the introduced pollen, but an appreciable part also came from introduced plants growing in low numbers and at more distant locations within the archipelago. Total long-distance transported pollen (35 taxa) increased from 3% of total pollen in the pre-human-impact period to 5% in the human-impact periods, probably due to destruction of native vegetation through fire and thus reduction of local pollen production. These phenomena might lead to erroneous interpretation of local plant occurrence when the native/non-native or local/extra-local status of plants is not known. © The Author(s) 2012.