1. Phosphorus (P) usually is the primary limiting nutrient of phytoplankton biomass, but attention towards nitrogen (N) and trace nutrients, such as iron (Fe), has surfaced. Additionally, N-fixing cyanobacterial blooms have been documented to occur in N-rich, P-poor waters, which is counterintuitive from the paradigm that low N and high P promotes blooms. For example, Lake Erie's central basin has Dolichospermum blooms when nitrate concentrations are high, which raises questions about which nutrient(s) are selecting for Dolichospermum over other phytoplankton and why an N-fixer is present in high N waters? 2. We conducted a 4-year (2014-2017) study in Lake Erie's central basin to determine which nutrient (P, N, or trace nutrients such as Fe, molybdenum [Mo], and boron [B]) constrained chlorophyll concentration, phytoplankton biovolume, and nitrate assimilation using nutrient enrichment bioassays. The enriched lake water was incubated in 1-L bottles in a growth chamber programmed at light and temperatures of in situ conditions for 4-7 days. We also quantified heterocytes when N-fixing cyanobacteria were present. 3. Compared to the non-enriched control, the P-enriched (+P) treatment had significantly higher chlorophyll and phytoplankton biovolume in c. 75% of experiments. Combination enrichments of P with ammonium-N, nitrate-N, Fe, Mo, and B were compared to the +P treatment to determine secondary limitations. +P and ammonium-N and +P nitrate-N resulted in higher chlorophyll in 50% of experiments but higher phytoplankton biovolume in only 25% of experiments. These results show that P was the primary limiting nutrient, but there were times when N was secondarily limiting. 4. Chlorophyll concentration indicated N secondary limitation in half of the experiments, but biovolume indicated only N secondary limitation in 25% of the experiments. To make robust conclusions from nutrient enrichment bioassays, both chlorophyll and phytoplankton biovolume should be measured. 5. The secondary effects of Fe, Mo, and B on chlorophyll were low (<26% of experiments), and no secondary effects were observed on phytoplankton biovolume | 2155 CHAFFIN et Al.