Rhododendron maximum (Rm) thickets suppress overstory regeneration in the southern Appalachians. We examined if Rm litter input and quality alters N cycling in hardwood forests. Post leaf-fall leaf litter biomass and fine root standing biomass were significantly greater under Rm than under hardwood stands, suggesting that Rm increases litter inputs. We assayed the potential for protein-tannin (P-T) complexing by fine root and leaf litter extracts from Rm and hardwood species. Tannins from Rm fine roots and leaf litter had a greater capacity to create P-T complexes compared to hardwood litter tannins. Since litter inputs are greater under Rm than under hardwoods, and Rm litter tannin has a greater capacity to precipitate protein, there is a greater potential for P-T complexing in Rm soils. Activity of soil polyphenol oxidases (PPO), enzymes that degrade P-T complexes, was significantly greater under Rm thickets compared to hardwood stands in the O horizons. From a reciprocal leaf litter enzyme study, PPO activity was also greater under Rm, regardless of leaf litter type, suggesting that the Rm soil microbial community can better access N in P-T complexes. In contrast, protease activities (enzymes that cleave proteins) were not significantly different between Rm thickets and hardwood stands. Protease production is more widespread across microbial flora compared to PPO enzyme production. Rm associates with a highly saprotrophic ericoid mycorrhizal fungal community with a documented ability to produce PPO enzymes, unlike the ecto- or arbuscular mycorrhizal fungi of hardwood species. Therefore, a positive-feedback between Rm litter quality and ericoid mycorrhizal fungi may allow Rm to tap into, while concurrently limiting hardwood species access to, a recalcitrant N pool. This feedback may partially explain why Rm so successfully excludes competing species.