Abstract: ‍ ‍Space targets， such as satellites， play a critical role in remote sensing mapping， meteorological monitoring， wireless communication， and reconnaissance， which serve as essential objects in space situational awareness. Polarimetric inverse synthetic aperture radar （ISAR） offers unique advantages for space target awareness by providing high-resolution images and polarimetric scattering information， which are sensitive to target structures. Nevertheless， challenges persist due to the scarcity of publicly available real measured data for polarimetric ISAR space targets and the complexity of fine-grained component-level annotation， which remains underexplored. To address these challenges， this study introduced the Pol-ISARSpaceTarget-1.0 dataset， the first publicly available dataset for the fine recognition of polarimetric ISAR space targets. The dataset consists of polarimetric ISAR images of six space targets， including radar satellites with parabolic antennas， two radar satellites with plane antennas， a communication satellite， and an optical satellite. It also contains the semantic annotations of nine typical components. This study details the dataset composition， electromagnetic simulation and imaging， and annotation workflow. To validate its effectiveness， representative deep-learning object-recognition methods were adopted for comparison studies， producing benchmark results that offer valuable references for researchers. The dataset features various kinds of space targets， comprehensive polarimetric information， and detailed component labels that provide fundamental data support for the fine classification and recognition of space targets.