Vegetation is a key component controlling soil accretion in coastal wetlands through production of belowground organic matter and enhanced deposition of mineral sediments. Vegetation structure is a proxy for wetland health and degradation that can be monitored at large scales with remote sensing. Among different multispectral indices, the Normalized Difference Vegetation Index (NDVI) is generally used for this purpose. Using Google Earth Engine (GEE), NDVI time‐series are extracted around 45 monitoring stations of the Coastwide Reference Monitoring System (CRMS) located in Terrebonne Bay, Louisiana, USA. NDVI tends to increase from saline to freshwater wetlands. Using these NDVI observations and in situ measurements of salinity, soil accretion rates, and geomorphic metrics (i.e., elevation, distance from the bay or from the nearest channel bank), empirical models were developed to derive maps of organic mass accumulation rates and salinity. The analysis shows that NDVI can be used to reproduce the salinity gradient in Terrebonne Bay, as the index captures differences in vegetation cover, which depend on salinity. A negative relationship between NDVI and organic accumulation mass rates is also found, indicating that saline marshes tend to accumulate more organic material compared to fresh wetlands.