Abstract: Ionic rare earth is an important strategic resource in China, and the southern Jiangxi province accounts for 2/3 of the total storage in the country. However, the ammonia nitrogen pollution problem which generated during the rare earth mining process is also need to be solved. This study was conducted to reveal the surface runoff production and residual ammonia migration characteristics of rare earth leaching collapsed debris deposits and provide a scientific basis for ammonia nitrogen pollution control in ionic rare earth mining areas. Three natural slope runoff plots were constructed based on typical ore leaching collapsed debris deposits in southern Jiangxi province. Rainfall runoff monitoring was carried out to analyze the surface runoff production and ammonia nitrogen migration characteristics of the collapsed debris deposits. The results show that the runoff coefficient of the collapsed debris deposits in the mining area varies from 0.01 to 0.3, the surface runoff depth and runoff coefficient can be characterized by the linear function of rainfall characteristic factor, and the fitting accuracy of the composite rainfall factor is better than that of the single factor. The rainfall × maximum 10-min rainfall intensity is the best method for predicting the runoff production on the collapsed slope(maximumR²=0.97).The average concentration of ammonia nitrogen in surface runoff is between 1.58 mg/L and 5.99mg/L, which is higher than the Surface Water Environmental Quality Standard (GB3838-2002) for Class Ⅴ water. Runoff ammonia concentration is negatively correlated with rainfall and rainfall duration, and positively correlated with other rainfall factors. The linear equation of maximum 30-min rainfall intensity has the best fitting accuracy for ammonia concentration(maximumR²=0.83). There is a risk of ammonia pollution for the surface runoff of the collapsed debris deposits of rare earth leaching ore in the study area, and the collapsed wall is the key part of residual ammonia nitrogen migration from the leaching mountain to the surface. The fitting equations based on the maximum 10-min and maximum 30-min rainfall intensity indexes can be used to predict surface runoff and runoff ammonia nitrogen, respectively. The above results could provide equations or critical parameters for related pollution risk evaluation models, therefore would be helpful in predicting ammonia pollution in the ionic rare earth mining areas of southern Jiangxi province. It could also be adopted in guiding conservation treatments for local soil and water environment.