Despite favorable conditions for enhanced biological phosphorus removal (EBPR), there are instances where good phosphorus (P) removal is elusive. Lack of performance may be due to either excess secondary P release or dominance of Glycogen Accumulating Organisms (GAOs) over Phosphate Accumulating Organisms (PAOs).Several wastewater treatment plants were identified to illustrate what is currently known about PAOs, GAOs, secondary release of P, and conditions favoring GAO growth over that of PAOs. It will be demonstrated how one can study nutrient profiles through a plant, manipulate data, and perform experiments to determine the cause of EBPR failure. In the plants studied, it was shown that successful BPR requires sufficient influent VFA, high primary P release in the anaerobic zone, no release but rather some uptake of P in the anoxic zone, and good P uptake in the aerobic zone. In an example where secondary release was found to be responsible for poor BPR performance, good primary release in the anaerobic zone was offset by substantial secondary release in the anoxic zone, likely caused by inadequate mixed liquor recycle with rapid depletion of nitrates in the anoxic zone. This combined with a high RAS recycle rate, which returned excessive nitrates and reduced the anaerobic retention time, could account for the high effluent P. GAO dominance could not be definitively proven in any of the cases studied; in cases where GAOs were named as a possible cause for unsuccessful BPR, influent VFA concentrations were also unknown and there is the possibility that low influent VFA could be to blame.Profiles of nutrient concentrations through BNR plants can be difficult to interpret due to the influence of recycle streams, but a mass balance through the plant will reveal the true state of P release and uptake. Either secondary release or GAO dominance could be the cause of inadequate P removal in the presence of sufficient influent VFA. If secondary release is not evident from the data and the release of P in the anaerobic zone appears inadequate, a test can be performed to indicate GAO dominance. Recorded instances of BPR failure at a full-scale plant due to GAO dominance are rare. While PAO/GAO population dynamics remain an important and ongoing area of study, it is critical that the presence of GAOs in activated sludge is not used as the only possible explanation for non-performance without first examining all possibilities for secondary release.Knowing how to identify cases of GAO dominance and secondary release of P by studying P profiles through the plant will allow operators to focus their energies on process adjustments or changes that will improve EBPR operation and performance. A full understanding of the phenomena that lead to EBPR failure, coupled with good design WEFTEC ® 2003 and operational practices, will allow for optimization of the EBPR system and consistent compliance with stringent effluent P permit requirements.