[Objective] Net primary production (NPP) of vegetation at the Dongtan wetland in Chongming District, Shanghai City was estimated, and methods for NPP estimation in small-scale wetland areas were determined, in order to provide data support for the effective assessment of carbon sequestration capacity in wetland ecosystems. [Methods] The study focused on Dongtan wetland, and used Sentinel-2B satellite remote sensing data and ground-based meteorological data. Monthly, seasonal, and annual-scale NPP estimates were made for 2017—2021 using the CASA model. Correlation analyses were performed to examine natural factors influencing the spatial-temporal variation in NPP. [Results] ① Mean NPP values (calculated by C) for the study area from 2017 to 2021 were 390.22±155.68, 426.74±102.40, 575.25±445.51, 539.29±201.87 and 611.81±464.88 g/(m²·yr), respectively, with total amounts (calculated by C) of 1.56×10¹⁰, 1.96×10¹⁰, 2.54×10¹⁰, 2.70×10¹⁰, and 2.94×10¹⁰ g/yr, showing a clear increasing trend. ② High NPP areas continuously expanded towards the sea, with Phragmites australis accounting for 54.73%—70.03% of the total NPP in the study area, followed by Spartina alterniflora and Scirpus mariqueter. ③ Monthly mean NPP values exhibited a normal distribution, and NPP values showed significant spatial variations with the seasons. Average NPP values and total NPP were notably higher in summer and autumn, with less pronounced changes in spring and winter. [Conclusion] Monthly average temperature was identified as the primary factor influencing monthly mean NPP values. The wetland vegetation classification results was obtained by using high spatial resolution remote sensing images， the authenticity of estimation results could be improved by localizing estimation model parameters.