We investigated whether marsh surface elevation, plant community composition (annuals vs. perennials), and organic matter quantity/quality were associated with differences in denitrification rates in an urban tidal freshwater marsh of the Potomac River, United States. We measured denitrification rates using both denitrification enzyme activity (DEA) with acetylene inhibition (June: n = 38, 3234 +/- 303; October: n = 38, 1557 +/- 368 ng N g dry soil(-1) h(-1)) and direct N(2) flux measurements with membrane inlet mass spectrometry (MIMS) (November: n = 6, 147 +/- 24 mumol m(-2) h(-1)). Organic carbon content and nitrate concentrations in soil, and plant community composition were correlated with elevation, but DEA rates did not differ across marsh surface elevation. Soil organic carbon was highest in plots dominated by perennial graminoids, but DEA rates did not differ across plant community types. The DEA rates increased with increasing soil ammonium concentrations and total N content, and DEA rates differed between summer and fall sampling. The MIMS rates did not differ across plant community types, but were correlated with soil organic N content. Denitrification rates suggest that potential N removal at the site could be substantial. In addition, denitrification rates measured in Dyke Marsh were higher than rates for sediments measured in the adjacent Potomac River. Tidal freshwater marshes can represent an important site for denitrification, and factors fostering denitrification should be considered when restoring urban tidal freshwater wetlands as they are faced with pressures from increasing land use change and sea level rise.