Polyphosphate (poly-P) is a major constituent in activated sludge from wastewater treatment plants with enhanced biological phosphorus removal due to poly-P synthesis by poly-P accumulating organisms where it plays an important role for recovery of phosphorus from waste water. Our aim was to develop a reliable protocol for poly-P quantification by 31 P NMR spectroscopy. This has so far been complicated by the risks of inefficient extraction and poly-P hydrolysis in the extracts. A protocol for complete extraction, identification and quantification of poly-P in activated sludge from a waste water treatment plant was identified based on test and evaluation of existing extraction protocols in combination with poly-P determination and quantification by solution and solid state 31 P NMR spectroscopy. The total poly-P middle group content was quantified by solid state NMR for comparison with the poly-P middle groups quantified by solution NMR, which is novel. Three different extraction protocols previously used in literature were compared: 1) a single 0.25 M NaOH-0.05 M EDTA extraction, 2) a 0.05 M EDTA pre-extraction followed by a 0.25 M NaOH main extraction and 3) a 0.05 M EDTA pre-extraction followed by a 0.25 M NaOH-0.05 M EDTA main extraction. The results showed that the extraction protocol 2 was optimal for fresh activated sludge, extracting 10.8 ± 0.4 to 11.4 ± 1.2 mgP/gDW poly-P. Extraction protocols 1 and 3 extracted less than 9.4 ± 0.5 mgP/gDW poly-P. A comparison of the quantification of poly-P by 31 P solution NMR and by 31 P solid state NMR spectroscopy of lyophilised activated sludge showed 86 ± 9% extraction efficiency of poly-P, which confirms that the extraction protocol recovered most of the poly-P from the samples without pronounced poly-P degradation.