<p>Soil erosion in forested watersheds poses significant environmental challenges by accelerating land degradation, reducing water quality, and altering ecosystem functions. Modeling soil erosion is vital to quantify erosion rates, design mitigation measures, and support land and watershed management policies. This systematic review was conducted to (i) assess the most frequently used soil erosion models in peer-reviewed literature, (ii) examine the temporal and spatial distribution of these models in forested watersheds, and (iii) evaluate the range of soil erosion rates estimated using different modeling approaches. A total of 105 peer-reviewed studies published between 1995 and 2021 were analyzed. The findings reveal a dominance of empirical models particularly the universal soil loss equation (USLE) and its revised version (RUSLE) which together account for over one-third of the applications. These models are widely applied due to their simplicity and minimal data requirements but may oversimplify the dynamic processes in forest environments. In contrast, process-based models such as WEPP offer greater simulation capacity but are underutilized due to calibration complexity and high data demands. The spatial distribution of studies is skewed toward Asia, North America, and Europe, with limited representation in Africa, South America, and Oceania. Temporal analysis shows a rising trend in publications since 2010. Therefore, this review offers a comprehensive overview of the current state of the art in soil erosion modeling and its applications globally, aiming to support soil erosion assessments and to inform research priorities by establishing a foundation for future targeted and in-depth analyses.</p>