Two Nitrification-Denitrification Biological Excess Phosphorus Removal (NDBEPR) systems have been operated for 8.5 months in order to compare their Biological Excess Phosphorus Removal (BEPR) performance. One of these systems, i.e. the University of Cape Town (UCT) system, exhibits mainly aerobic P uptake while the External Nitrification Biological Nutrient Removal Activated Sludge (ENBNRAS) system is characterised by high anoxic P uptake. It was observed that when operating with predominantly aerobic P uptake, the UCT system released more P than the ENBNRAS system, even though it had a lower anaerobic mass fraction. However, when the influent TKN/COD was high, i.e. > 0.1, anoxic P uptake also occurred in the UCT system and P release dropped to lower levels than in the ENBNRAS. Accordingly, P uptake of the UCT system was 5 mg P/l influent higher than that of the ENBNRAS system, when it was predominantly aerobic, but 9 mg P/l influent lower when anoxic P uptake occurred. As a result, the UCT system achieved superior P removal when aerobic P uptake was predominant (23% higher), but when high influent TKN/COD promoted anoxic P uptake the P removal of the UCT system was poorer than that of the ENBNRAS system. This study clearly showed that anoxic P uptake is not beneficial to NDBEPR systems.