The canopy water content of wetland vegetation is an important measuring index of the health status of wetland ecosystem. This article takes the Honghe national wetland nature reserve as study area. We focus on innovative approaches for retrieving canopy water content from optical remote sensing data-multispectral and hyperspectral data. Spectral features, such as narrow band spectral indices, hyperspectral vegetation indices in early literatures, absorption features and vegetation indices extracted from TM image were used to estimate the canopy water content. For narrow band spectral indices, Normalized difference vegetation index comprised of 970 nm and at 900 nm had a highest correlation with canopy water content. For general hyperspectral vegetation indices in early literatures, WI had a highest correlation with canopy water content. For absorption features, the absorption deepness at 1200nm had a highest correlation with canopy water content. In addition, NDII (band5) extracted from TM images could be used for estimating canopy water content. Finally, a distribution map of canopy water content in HNNR was generated.