Here we report the results of a five-year shallow-lake mesocosm experiment testing the role of nitrogen (N) in inducing and reversing lake eutrophication. We tested two conditions: 1) the effect of combined but imbalanced N and P fertilization on inducing lake eutrophication, and 2) the effect of subsequently eliminating N fertilization while maintaining P fertilization to reverse lake eutrophication. The mesocosms were open to both the atmosphere and the lake sediments which ensured that the experimental conditions mimicked the complex biogeochemical interactions that occur in lakes. We fertilized all mesocosms with 1.3 g P m-2 y-1 for five years. We also fertilized with four different N rates (1.3, 9.4, 32, and 64 g N m-2 y-1) in a randomized block design and utilized the lake outside the mesocosms as a control. After four years, we halted all N fertilization but continued P fertilization for a fifth year. Nutrient concentrations and phytoplankton biomass responded approximately proportion to N and P fertilization rates during the first four years of the experiment. Total N concentrations declined rapidly after the cessation of N fertilization but total P concentrations actually increased despite constant P fertilization. Chlorophyll-a concentrations decreased by 50-70% and was proportional to total N concentration reductions. Combining our results with the 2022 National Lakes Assessment data we show that reducing N inputs has the potential to reverse eutrophication in over 1 million lakes across the conterminous US.