Indigenous cultures know a great deal about the landscape they inhabit, and their knowledge can be a valuable tool for ecologists. In order to explore how residents' knowledge might help characterize a large and diverse forest type in southeastern Peru, we asked plant experts of the local Cashinahua culture to predict whether the tree species recorded in a single 1-ha plot in upland forest were common on the surrounding landscape. We then compared their answers with data collected in four other 1-ha plots scattered over an area of about 7,000 km 2 . Cashinahua predictions matched tree plot data for 66% of the species examined. Species labeled as common by the Cashinahua included 9 of the top 11 most common species in the 5 plots and 39% of all trees in the plots. We discuss three obstacles to using local knowledge in large-scale vegetation studies: 1) the often-confusing relation between indigenous and Linnaean taxonomic nomenclature, 2) differing cultural conceptions of commonness and rarity, and 3) the limitations of describing tree species abundance via 1-ha tree plots. Where these limitations can be overcome, studies of large-scale vegetation patterns stand to benefit greatly from incorporating local knowledge of regionally abundant species.