Abstract: Abstract: Improvement of area accuracy and reliability is highly important for the dynamic monitoring of mangrove forests based on remote sensing. In this study, two serious problems present in certain studies addressing mangrove dynamics were analyzed: one was that in the same region and same year, the mangrove areas reported by different researchers varied greatly, and the other was that the area of mangrove forests in a region or gulf changed dramatically in a short period. The main factors affecting the accuracy of mangrove area surveys using remote-sensing data at regional and local levels were discussed based on the practical experience of monitoring of mangrove dynamics from many years. The spatial resolution of remote-sensing images, tidal height, protocol for extraction of information on mangrove distribution, and field validation were the most important factors affecting the accuracy of mangrove areas based on remote sensing. We suggested that remote-sensing data of extremely high resolution(1 m or below)should be used to extract information of mangrove areas at regional and local levels in China; tides were low, and the beach was exposed when the image was acquired. When remote-sensing information from two or more periods is extracted, the area of mangrove patches should be prevented from changing dramatically between adjacent periods. Furthermore, a basic prerequisite for any image-analysis approach is the conduction of an intensive field study; the characteristics of mangrove distribution in each gulf or bay should be investigated by field work, and then, the patch boundary should be revised individually in the field.