Landslides caused by underground mining are very common in mountain mining areas, and thus threaten human lives and assets. Under the superposed influence of underground mining and landslides, surface movement in mountain areas is different from that in flat areas. Further understanding of the law of surface movement in mountain mining areas can protect the locals and their assets. In this study, an improved survey technique called spherical fitting real-time kinematic was employed to monitor surface movement regularly, and similar material simulation was used as a supplement to reveal the process of strata failure. Results showed that the ground would uplift during underground mining in the valley bottom areas, an action that caused the subsidence velocity curve to appear as a skewed distribution instead of a Gaussian distribution. To improve the prediction accuracy, the prediction model of surface subsidence in mountain mining areas was updated. The mountain surface subsidence prediction system was developed using Visual Basic. Finally, the mechanism of mining landslide was studied. This work is of general interest to the professional public and it will contribute to a more objective knowledge of understanding the surface movement in mountain mining areas.