Phosphorus-31 (31P) nuclear magnetic resonance (NMR) spectroscopy of sodium hydroxide–ethylenediaminetetra-acetic acid (NaOH-EDTA) extracts has recently become a widely used technique for the characterization of soil P. This technique has seemingly enabled easy identification and quantification of phytate (myo-inositol hexakisphosphate), a compound long believed to constitute a major proportion of organic P. Phytate is usually identified by its characteristic pattern of four resonances in the ratio 1:2:2:1. We report that the 31P NMR spectra of the NaOH-EDTA extracts of four Australian pasture soils contain a set of resonances that bear a striking resemblance to the phytate resonances but that are shown not to be phytate though careful addition (spiking) of pure phytate. The spiking experiments identify a much smaller set of resonances as being phytate. Quantification of these resonances shows that phytate comprises <5% of organic P and <3% of total P in these soils. We also show that the 31P chemical shift of phytate resonances is very sensitive to pH and ionic strength. These results highlight the potential for misassignment of resonances in 31P NMR spectra of NaOH-EDTA extracts of soil and the possibility that phytate concentrations may be overestimated using this technique and show the value of spiking as a definitive form of identification of 31P NMR resonances.