Pretreatment with chemicals such as ethylenediaminetetraacetic acid (EDTA) is often used to improve the analysis of sediment P with solution P-31 nuclear magnetic resonance spectroscopy (35P-NMR), but there is a lack of a comprehensive assessment of the methodology. In this study, the effects of EDTA pretreatment on sediment P extracted using a mixture of 0.25 mol L(-1) NaOH and 50 mmol L(-1) EDTA (NaOH-EDTA) were examined with 45 different sediments. The results showed that EDTA pre-extraction decreased the amount of P extracted by NaOH-EDTA when the concentration ratio of sediment Ca to the sum of sediment Fe and Al [Ca/(Fe+Al), on a wt/vol basis] was lower than 0.4. An increase in total extracted P, coupled with substantial increases in total extracted paramagnetic ions such as Fe and Mn, was observed for another group of sediments with Ca/(Fe+Al) > 0.5, possibly due to the matrix effect. Analysis of 16 representative sediments with 31P-NMR showed that orthophosphate diesters were substantially removed by EDTA pre-extraction for sediments with Ca/ (Fe+Al) between 0.4 and 0.7, reflecting a high risk posed by this pretreatment. Phosphorus diversity and concentration of individual P compounds were markedly improved for sediments with Ca/(Fe+Al) > 0.7, suggesting that EDTA pretreatment was particularly useful for 31P-NMR analysis of calcareous sediments. The present study showed that sediment properties played an important role in determining pretreatment effects. Caution is advised when applying pretreatment methods to different sediments.