Irradiation of reactor pressure vessel (RPV) steels causes the formation of nanoscale microstructural features (termed radiation damage), which affect the mechanical properties of the vessel. A key tool for characterizing these nanoscale features is atom probe tomography (APT), due to its high spatial resolution and the ability to identify different chemical species in three dimensions. Microstructural observations using APT can underpin development of a mechanistic understanding of defect formation. However, with atom probe analyses there are currently multiple methods for analyzing the data. This can result in inconsistencies between results obtained from different researchers and unnecessary scatter when combining data from multiple sources. This makes interpretation of results more complex and calibration of radiation damage models challenging. In this work simulations of a range of different microstructures are used to directly compare different cluster analysis algorithms and identify their strengths and weaknesses.