An Annual nitrogen budget is presented for a small steam draining Watershed 10, H. J. Andrews Experimental Forest, Oregon. The role of allochthonous debris in the input, flux, and export of nitrogen is emphasized in the material balance budget. All material entering the stream channel was presumed to enter the water sometime during the year. Material estimates are based on total channel area. The major annual nitrogen input (1974—1975) was subsurface flow (11.06 g/m2) as dissolved organic nitrogen (10.56 g/m2) and nitrate (0.50 g/m2).Biological inputs of nitrogen amounted to 4.19 g/m2as direct terrestrial inputs of: litterfall (1.35 g/m2), lateral movement (1.78 g/m2), and throughfall (0.30 g/m2). Nitrogen fixation on fine wood debris contributed an additional 0.76 g/m2based on rates from a nearby watershed. Total nitrogen input was 15.25 g/m2. The nitrogen pool was dominated by large amounts of particulate organic matter. Coarse wood constituted 32% of the nitrogen pool (3.80 g/m2) and fine wood fractions 18% (2.18 g/m2). The coarse wood fraction greatly influenced stream morphology. Fine organic particulates constituted an additional 40% of the nitrogen pool (4.77 g/m2). DON (dissolved organic nitrogen) export (8.38 g/m2) was less than input, presumably due to biological uptake associated with litter mineralization, sorption, and chemical flocculation. Due to effective retention of particulate inputs by debris dams, biological processing in the particulate nitrogen pool, and uptake and sorption of DON, most particulate organic inputs increased in nitrogen concentration prior to export. Particulate organic nitrogen input (3.13 g/m2) was greater than export (2.53 g/m2). Total annual nitrogen output was 11.36 g/m2, resulting in a gain of 3.89 g°m—2°yr—1to the stream. Thus, the stream was not operating on an annual steady state, but on an input—output regime related to the processing of refractory wood debris and resetting by major storms. Although particulate and dissolved nitrogen loss per hectare was small for the 10—ha watershed, these losses passed through or were accumulated in a pool encompassing<1% of the watershed area. This concentration of N in the stream allowed establishment of a separate ecosystem whose processing efficiency and capabilities for nutrient cycling were related to the retention capacity of the channel and nutrient quality of inputs within the reach.