Nutrient pollution is degrading global lake water quality. Therefore, it is critical to understand how lakes will respond to interventions aimed at reversing eutrophication. Historically, eutrophication research focused on nutrient addition experiments, pointing to phosphorus (P) as the key driver of nuisance algae. Management efforts have successfully reduced P inputs and availability in many lakes, but phytoplankton biomass shows variable responses to nutrient control efforts. This suggests that assuming nutrient removal will induce the opposite effect of nutrient addition ignores important ecological dynamics. I predict that phytoplankton physiology and community composition will adjust after P removal to favor better P competitors. This will result in weaker P-limitation and a smaller reduction in phytoplankton biomass than anticipated. To characterize the multi-year nutrient limitation response to P removal, I leveraged whole-lake, nutrient removal experiments (alum treatments) in Minnesota, USA. I sampled a chronosequence of 14 shallow lakes that were alum treated in the last 5 years, along with 4 untreated control lakes. Nutrient limitation was determined from a factorial amendment of nitrogen (N) and P on incubated lake water. Phytoplankton communities were characterized using 16S and 18S amplicon sequencing. Results suggest that P-limitation weakens each year after treatment, to the point where many lakes become co-limited by N and P. This shift is partially, but not entirely, explained by a rebound in surface water P. The phytoplankton community also changes in the years following treatment, which may help explain the existence of co-limitation and P-limitation at similar N:P. Further research is needed to untangle the relative importance of changes in community composition versus phytoplankton physiology. This study shows that P removal only temporarily induces P-limitation. The subsequent shift to co-limitation indicates that the phytoplankton community has come into balance with the new nutrient regime, explaining why reductions in P are not always accompanied by sustained reductions in phytoplankton biomass.