BorFIT is a comprehensive training dataset that aims at enhancing the segmentation of individual trees and species detection from LiDAR point clouds, specifically in boreal forests. This dataset comprises 385 LiDAR point clouds, each covering an area of 20×20 m² and containing up to 200 manually segmented trees. The data was collected from 145 sites across Yakutia, Canada, and Alaska between 2021 and 2023, strategically chosen to represent a bioclimatic gradient across the circumboreal region. The LiDAR surveys utilized a YellowScan LiDAR Mapper+ mounted on a UAV which surveyed the areas at an altitude of 70 m achieving an average point density of 1200 points/m² over an area of approximately 50×500 m². Up to 4 reference plots (depending on data quality and performed forest inventories) per point cloud were extracted, based on present tree heights to generate a representative subset of the original point cloud. Manual segmentation of trees was performed using the software CloudCompare. Based on these, a training data set for a randomForest classifier was prepared by assigning the species determination for a subset of individual trees per present species. The classifier was then trained based on structural parameters from the point cloud, and if available, spectral information based on RGB imagery (CA,AK). The classifier was then utilized to predict the species of all segmented trees within the data set. The number coded species prediction was added to a new scalar field in the point clouds including a probability value as a indicator for prediction accuracy. The dataset encompasses various mostly northern boreal tree species, including Picea glauca, Picea mariana, Betula spp., and Larix laricina, among others. This diversity supports the analysis of species distribution and stand structure, crucial for understanding vegetation dynamics in response to climate change. The dataset's design not only addresses the limitations of existing AI-supported detection methods but also serves as a foundational resource for future research into boreal forest dynamics under global warming scenarios.