Wetlands,altogether with forests and oceans,listed as the worlds three major ecosystems,are unique ecosystems resulted by the interaction of land and water on earth with huge resource potential and environmental,social,and economic functions.Wetlands not only provide a variety of resources for human living and production,but also play an irreplaceable role on against floods,runoff regulation,environmental improvement,pollution control,diversity of species,gene and landscaping and maintenance of the regional ecological balance,gaining the reputation of “the kidney of earth” and “the cradle of life”.Henan province(31°23′N ~ 36°22′N,110°21′E ~ 116°39′E),China,locates in the middle and lower reaches of the Yellow River,where there are 4 major water systems of Yellow River,Huaihe,Yangtze river and Haihe with lots of wetlands around mountainous area,uplands and plains,especially the south of Henan,which belong to typical inland wetlands,divided into 5 types of river,lake,swamp,beach and paddy field.Remote sensing (RS) technology is an effective tool for wetland resource investigation.Traditional field sampling methods take a long time but have small coverage.In the wetland resource investigation,RS technology can provide mass data,which can reflect the internal wetland environment and study in-depth the relationship between wetland and climate,topography,vegetation,land use,biodiversity and social and economic developments after processing and analysis.In this paper,we extract the information of wetlands in Henan province using Landsat data in 3 periods of 1990,2000,2007 and generate database of wetland resources and make a systematic analysis of distribution,area,formation and change so as to provide scientific basis for wetland protection,retrieval and sustainable utilization.According to the test,minimum distance classification is fit for river features,parallelepiped classification for lake and reservoir and maximum likelihood classification for paddy fields.Beaches and swamps are classified by visual interpretation method because it is difficult to carry out by computer automatic classification.In the classification of lakes and reservoirs,mountain and cloud shadows are often misclassified.After analysis,these shadows can be rejected according to the different DN values of objects.Meanwhile,retrieval accuracy is improved by image fusion and spatial subset.Following conclusions can be obtained: a.artificial wetlands grow at a high speed.Paddy fields in Henan reach to 642600 hectares in 2007 from 518700 hectares in 1990,rising 23.87%.Meanwhile,a lot of reservoirs,canals and other water infrastructure are built.b.natural wetlands decline sharply.Swamps have the biggest change from 10680 hectares in 1990 dropped to 70 hectares in 2007,almost disappeared.Beachs dropped to 87500 hectares in 2007 from 200100 hectare in 1990.According to field survey,a lot of natural ponds dried or even disappeared.c.human factors are the main reason for the decreasing of natural wetland.According to satellite remote sensing and field survey,agricultural developments lead to the decrease of swamp.The Yellow River Gudao wetland gradually disappeared because of the construction of Xiaolangdi reservoir.Otherwise,city constructions also take up a lot of beachs.