AbstractOur primary objective was to compile a comprehensive dataset on bird–window collisions throughout the Neotropical region, including both published and unpublished sources. On May 12, 2020, we extensively disseminated invitations to provide data via email and social media platforms. By providing a template worksheet, we required standardized information from collaborators to complete and register their data. To better understand how these data were acquired (e.g., incidental observations and systematic procedures), we sent out a survey to all collaborators. We established rigorous validation criteria for data inclusion and conducted thorough curation procedures to ensure accuracy. After the filtering process, we compiled a total of 4103 bird–window collision reports. These came from 11 Neotropical countries, dating from 1946 to 2020, and revealing distinct regional patterns and potential seasonal patterns. The five most frequent orders were Passeriformes (2451), Columbiformes (520), Apodiformes (377), Psittaciformes (202), and Piciformes (186). Data on bird–window collisions were collected through a local specific systematic protocol (1419), by chance (1252), by government agencies (742), and by other approaches (632), while a few reports were collected by unknown procedures (58). The volume of records across months in our dataset suggests that there may be temporal patterns, with peaks: the first one in March–April and the second one in October–November, which seem to align with the major migration and reproduction seasons. This dataset represents the first comprehensive effort in the Neotropical region focused on bird–window collision data, providing valuable insights for further scientific advancements and conservation policies. The data are free from copyright or proprietary restrictions. Please cite this data paper when using the data in publications or scientific presentations.