Abstract: The establishment of a scientific and systematic indicator system for “green watersheds” is a fundamental prerequisite for achieving sustainable watershed development. This study systematically integrated multi-objective and multi-dimensional indicators from domestic and international watershed evaluation studies using meta-analysis. A comprehensive indicator set for green watersheds was subsequently compiled based on a “Functionality-Integrity-Systemicity” framework. Through correlation analysis and heterogeneity tests, this study revealed the degree of correlation and heterogeneity characteristics between indicators from the three dimensions and the overall green watershed level. The results show that within the Functionality dimension, natural capital indicators such as the vegetation index (R = 0.75) and forest coverage rate (R = 0.75) exhibited strong positive correlations with the green watershed level, whereas soil erosion (R = −0.47) and the soil and water loss rate (R = −0.50) showed significant negative correlations. In the Integrity dimension, the important habitat retention rate (R = 0.71) and the biodiversity index (R = 0.60) demonstrated consistently high correlations, highlighting the pivotal role of ecological indicators. For the Systemicity dimension, a positive association was found between GDP per capita (R = 0.64), urbanization level (R = 0.62), and the green watershed level. Furthermore, subgroup analysis was conducted to preliminarily investigate the sources of high heterogeneity for specific indicators. Based on these findings, the study discusses key challenges in green watershed evaluation, including differentiated assessment, the scale dependency of indicator systems, and the construction of dynamic evaluation frameworks. Ultimately, this research provides a scientific reference for the development of differentiated and precise green watershed evaluations.