The majority of marine nitrogen loss occurs not in the water column but in the sediments. Contrasting to anammox bacteria which requires access to both oxidized (NH₄⁺) and reduced nitrogen (NO₂^-) at the same time, denitrifying bacteria theoretically could occupy a wider range of redox conditions since their substrates (organic matter and nitrate) largely co-exist. However, where and how denitrification happens in the vast habitat remain unclear. From the microbial physiological perspective, it seems require oxygen-deficient zones because the key enzymes of denitrification are sensitive and can be severely inhibited by oxygen. Here, we argue that denitrification mainly occurs in the surface sediments despite the presence of oxygen, due to the following reasons. (i) Measured nitrate concentrations are lower than those estimated from the nitrate production by nitrification and also from the stoichiometrically calculation from the measured phosphate concentrations; and (ii) functional genes of denitrification are most abundant in the surface sediments. We also reconstructed metagenome-assembled genomes (MAGs) of novel archaea and bacteria that are putatively involved in denitrification in marine sediments, which are distinct from prominent denitrifiers in seawater. Ongoing work includes the dual nitrogen and isotope analyses of porewater nitrate to provide more in situ evidence of denitrification in oxic sediments. The combination of these complementary data illuminate organisms and their activity in one of the most important components of nitrogen cycling in the global ocean.