Tidal wetlands play an important role in global carbon cycling by storing carbon in sediment at millennial time scales, transporting dissolved carbon into coastal waters, and contributing significantly to global CH4 budgets. However, these ecosystems' greenhouse gas monitoring and predictions are challenging due to spatial heterogeneity and tidal flooding. We utilized eddy covariance and chamber measurements to quantify fluxes of CO2 and CH4 at a restored tidal saltmarsh across spatial and temporal scales. Eddy covariance data revealed that the site was a strong net sink for CO2 (−387 g C‐CO2 m−2 yr−1, SD = 46) and a small net source of CH4 (0.7 g C‐CH4 m−2 yr−1, SD = 0.4). After partitioning net ecosystem exchange of CO2 into gross primary production and ecosystem respiration, we found that high net uptake of CO2 was due to low respiration emissions rather than high photosynthetic rates. We also found that respiration rates varied between land covers with increased respiration in mudflats compared to vegetated areas. Daytime soil chamber measurements revealed that the greatest CO2 emission was from higher elevation mudflat soils (0.5 μmol m−2s−1, SE = 1.3) and CH4 emission was greatest from lower elevation Spartina foliosa soils (1.6 nmol m−2s−1, SD = 8.2). Overall, these results highlight the importance of the relationships between wetland plant community and elevation, and inundation for CO2 and CH4 fluxes. Future research should include the use of high‐resolution imagery, automated chambers, and a focus on quantifying carbon exported in tidal waters. Plain Language Summary At the ecosystem level, a restored tidal salt marsh in the South San Francisco Bay California took in more carbon dioxide (CO2) from the atmosphere through photosynthetic activity than it emitted through respiration, and it emitted very small amounts of methane (CH4). This site appears to be a stronger sink for CO2 compared to other tidal marsh sites due to the very low rate of CO2 being lost through respiration to the atmosphere, rather than strong photosynthetic rates. We also found that ecosystem level CO2 emissions and the responses to temperature and light varied based on land cover type. By measuring soil surface emissions from each of the main land cover types of pickleweed, cordgrass, and mudflats we found that on average soils with lower elevation where cordgrass grows were stronger sources of CH4 while mudflat soils with greater elevation were stronger sources of CO2. Key Points Soil chamber measurements were able to detect significant differences in CO2 and CH4 fluxes between land cover types Vegetation and microtopography are drivers of the spatially heterogeneous CO2 and CH4 emissions within the wetland At the ecosystem level, high net uptake of CO2 was the result of low respiration emissions, suggesting lateral transport of dissolved CO2