To protect and manage Pacific Northwest salmonid stocks, sound management programs must be developed that incorporate knowledge of the ecological processes critical to their survival. One such process is the nutrient dynamic in headwater spawning streams. It is essential that we understand the trophic relationships among the spawners, the I to 3 year-old juveniles, the invertebrates, and the primary producers in these streams. The hypothesis that I tested is that photoautotrophic production increases due to the decomposition of salmon carcasses. Increased primary production would stimulate the production of invertebrates which are an important food resource of juvenile salmonid species. The hypothesis was tested by comparing periphyton communities sampled from upstream reference sites with those sampled from sites within and downstream of a onemile reach of Still Creek, a stream within the Zigzag Ranger District of the Mount Hood National Forest, in which approximately 300 coho (Oncorhynchus kisutch) carcasses were distributed. Comparisons were based on the numbers of each diatom species observed and periphyton biomass measured as chlorophyll a concentration and mass of organic matter lost on ignition. To more accurately characterize the response of the periphyton community, base-line data were collected before enrichment ascertaining the initial degree of nutrient limitation and measuring the variation within and among sample sites prior to site modification. Nutrient concentration and physical characteristics of the water were also monitored. The data collected indicate that the periphyton communities at each site were similar before carcass enrichment. After enrichment, treatment sites showed an increase in periphyton biomass. No change in species composition was observed from reference to treatment sites during the enrichment. Some suggestions are made for improving the experimental design to better characterize the response of the periphyton community to salmon carcass decomposition, such as increasing sample size, adding more "before treatment" sample times, and adding a second treatment stream.