We studied marsh-atmosphere exchange of carbon dioxide in a high marsh dominated salt marsh during the months of May to October in 2012-2014. Tidal inundation at the site occurred only during biweekly spring tides, during which we observed a reduction in fluxes during day and night. We estimated net ecosystem exchange (NEE), gross primary production (GPP), and ecosystem respiration (R eco ) using a modified PLIRTLE model, which requires photosynthetically active radiation, temperature, and normalized difference vegetation index (NDVI) as control variables. NDVI decreased during inundation, when the marsh canopy was submerged. Two-time series of NDVI, including and excluding effects of tidal inundation, allowed us to quantify the flux reduction during inundation. The effect of the flux reduction was small (2-4%) at our site, but is likely higher for marshes at a lower elevation. From May to October, GPP averaged −863 g C m −2 , R eco averaged 591 g C m −2 , and NEE averaged −291 g C m −2 . In 2012, which was an exceptionally warm year, we observed an early start of net carbon uptake but higher respiration than in 2013 and 2014 due to higher-air temperature in August. This resulted in the lowest NEE during the study period (−255.9 ± 6.9 g C m −2 ). The highest seasonal net uptake (−336.5 ± 6.3 g C m −2 ) was observed in 2013, which was linked to higher rainfall and temperature in July. Mean sea level was very similar during all 3 years which allowed us to isolate the importance of climatic factors.