Abstract: Site classification is one of the key topics in earthquake engineering, as it plays a crucial role in scientifically evaluating the influence of site conditions on ground motion and seismic damage, thereby providing a rational basis for determining seismic parameters and ensuring the seismic safety of engineering sites. Although there is a global consensus on the concept and theory of adjusting ground motion parameters according to different site types, significant differences still exist in the principles and methods used for site classification. In this study, the latest research progress on seismic site classification at home and abroad is comprehensively reviewed. From the perspectives of engineering geological conditions, inversion of ground motion records, and theoretical analysis, various classification methods and their underlying principles are systematically discussed. The internal relationships, distinctions, advantages, and limitations among different classification schemes are summarized in detail. Moreover, unresolved scientific issues in current site classification research are analyzed. In light of the rapid development and application of artificial intelligence (AI) technologies, the feasibility of incorporating AI approaches into site classification studies is proposed. Finally, key research directions and scientific challenges for future site classification in earthquake engineering are outlined. Through a comprehensive review of site classification research at home and abroad, the aim is to deepen the systematic understanding of seismic engineering site classification research, and provide reference and support for the revision of seismic design specifications, engineering seismic practice, and related scientific research work.