The analysis and prediction of the coastal erosion status are of great significance for maintaining the marine ecological environment, planning urban construction, and coping with climate change. Fangchenggang is located in the southwesternmost part of mainland China. It has the largest harbor in the western region, which provides access to the most convenient channel between southwest China and the Association of Southeast Asian Nations. With the rapid development of China’s economy, coastal erosion has occurred in Fangchenggang, and ecological and environmental problems have become increasingly prominent. In this study, we mainly focus on the status analysis and prediction of Fangchenggang coastal erosion based on multisource remote sensing data. A systematic analysis and prediction methods for the coastal erosion state are studied. First, shoreline type changes are analyzed. The index of coastline type diversity is used instead of solely using length changes to analyze shoreline diversity, and the length change intensity is applied to quantify the degree of shoreline change over time for different types of shorelines. Next, the composite index of coastline utilization degree is introduced to further explore the impact of human activities on shorelines. Then, to study the effect of coastal erosion on different types of shorelines in Fangchenggang, the end point rate is used to obtain the coastal erosion rate. Finally, the shoreline trend prediction model for Fangchenggang is established through the spatial distribution of the coastal erosion rate, and we verified that the accuracy of the model is 82%. The model can provide technical support for the integrated prevention and control of coastal erosion hazards.