Rainfall simulators are experimental devices that reproduce the essential characteristics of natural rainfall under controlled conditions. They serve as indispensable tools for decoupling rainfall characteristics from surface responses and have been widely used in soil science, hydrology, geomorphology, and environmental research. However, their technological development has long been constrained by the lack of standardization. This paper reviews the technological progression of rainfall simulators from early heterogeneous designs to more sophisticated systems, while also analyzing the types, characteristics, and comparisons of different rainfall simulators, as well as their applications in the fields of hydrology and soil erosion. Subsequently, the article systematically examines the limitations and challenges in rainfall simulator research. The discussion focuses on discrepancies in raindrop characteristics, spatiotemporal differences between simulated and natural rainfall, scale effects and uncertainties when extrapolating small-scale experimental results to watershed or regional scales, and on the uncertainty in discrepancy sources together with the associated challenges of standardization. Finally, it outlines future prospects for rainfall simulator development. Future development of rainfall simulators should strengthen systematic evaluation of rainfall characteristics and data quality. Integrating artificial intelligence can significantly enhance the accuracy and reliability of simulated rainfall analysis by capturing complex hydrological relationships and optimizing data processing. Meanwhile, conducting comparative analyses of different rainfall simulators will help identify configurations that most effectively replicate specific natural rainfall events. Moreover, developing and sharing a comprehensive rainfall simulator database will facilitate the integration of data from different simulators and enable performance evaluations, improving the accuracy and consistency of rainfall simulation results. Additionally, multiscale validation of the spatial representativeness of the data generated by rainfall simulators is essential. This will enhance the applicability of erosion model parameters across diverse landscapes. Lastly, addressing standardization challenges by establishing a limited comparability framework in specific fields and research contexts will harmonize implementation pathways and enhance the reliability and comparability of rainfall simulator data. This study contributes to advancing the development of rainfall simulators and provides a foundation for improving soil erosion and hydrological research.