Inorganic nutrient amendments to streams are viewed as possible restoration strategies for re-establishing nutrients and stream productivity throughout the western coast of North America, where salmon runs and associated marine-derived nutrient subsidies have declined. In a mesocosm experiment, we examined the short-term (6 weeks) comparative effects of artificial nutrient pellets and salmon carcasses, alone (low and high amounts) and in combination, on stream food webs. Response variables included dissolved nutrient concentrations, biofilm ash-free dry mass (AFDM) and chlorophyll-alevels, macroinvertebrate density, growth and body condition of juvenile coho salmon Oncorhynchus kisutch, and whole-body lipid content of invertebrates and juvenile coho salmon. Most of the response variables were significantly influenced by carcass treatment; the only response variable significantly influenced by fertilizer pellet treatment was soluble reactive phosphorus (SRP) concentration. Ammonium-nitrogen concentration was the only response variable affected by both (low and high) levels of carcass treatment; all others showed no significant response to the two carcass treatment levels. Significant treatment × time interactions were observed for all responses except nitrate; for most responses, significant treatment effects were detected at certain time periods and not others. For example, significantly higher SRP concentrations were recorded earlier in the experiment, whereas significant fish responses were observed later. These results provide evidence that inorganic nutrient additions do not have the same ecological effects in streams as do salmon carcasses, potentially because inorganic nutrient additions lack carbon-based biochemicals and macromolecules that are sequestered directly or indirectly by consumers. Salmon carcasses, preferably deposited naturally during spawning migrations, appear to be far superior to inorganic nutrient amendments for sustaining and restoring stream productivity, including fish production, and should be chosen over artificial nutrient additions when feasible and practical.