From December 1993 through November 1994, we conducted a nitrogen mass-balance study on a 0.74-ha, surface-flow wetland that receives water from an oxidation pond in the municipal waslewater treatment system in Arcata, California, USA. Water samples were collected biweekly and analyzed for concentrations of NO~-N (nitrate), NH,-N (total anamonia), and TN (total nitrogen). Flow was calculated from hydraulic head measurements taken at two inlet weirs. Macrophytes and sediment were sampled during summer to estimate N storage in these compartments. The wetland removed 17% of the 16,600 kg of influent N during the annual cycle. Little removal occurred prior to mid-May; whereas from mid-May through November, 37% ofinfluent N was removed. Expressed on an areal basis, net N removal was 1.84 g N.m 2-d during the warmer months of the year and 1.07 g N.m-:-d -~ during the entire 12-month period. Nitrogen storage in the seasonal, above-ground increment of macrophyte biomass was only 7% of overall N removal by the marsh. From 1986 to 1994, 990-2,390 kg of N (or 0.046-0.11 g N-m 2-d ~) accumulated in the sediment--only 4-10% of overall annual N removal. Mean annual TN removal (1.07 g N.m ~.d ') considerably exceeded mean influent , and removal of TN was highly correlated with NO,-N mass loading (r = 0.89, P < 0.001). Thus. sequential nitrification/denitrification appears to occur in lhe marsh, and denitrification (estimated at 0,95-1.02 g N-m -"-d -~) appears to be the primary N-removal mechanism,