Phosphorus-31 nuclear magnetic resonance (NMR) spectroscopy is an excellent tool with which to study soil organic P, allowing quantitative, comparative analysis of P forms. However, for 31P NMR to be tative, all peaks must be completely visible, and in their correct relative proportions. There must be no line broadening, and adequate delay times must be used to avoid saturation of peaks. The objective of this study was to examine the effects of extractants on delay times and peak saturation. Two samples (a forest litter and a mineral soil sample) and three extractants (0.25 M NaOH, NaOH plus Chelex (Bio-Rad Laboratories, Hercules, CA), and NaOH plus EDTA) were used to determine the differences in the concentration of P and cations solubilized by each extractant, and to measure spin-lattice (T1) relaxation times of P peaks in each extract. For both soil and litter, NaOH-Chelex extracted the lowest concentrations of P. For the litter sample, T1 values were short for all extractants due to the high Fe concentration remaining after extraction. For the soil sample, there were noticeable differences among the extractants. The NaOH-Chelex sample had less Fe and Mn remaining in solution after extraction than the other extractants, and the longest delay times used in the study, 6.4 s, were not long enough for quantitative analysis. Delay times of 1.5 to 2 s for the NaOH and NaOH-EDTA were adequate. Line broadening was highest in the NaOH extracts, which had the highest concentration of Fe. On the basis of these results, recommendations for future analyses of soil and litter samples by solution 31P NMR spectroscopy include: careful selection of an extractant; measurement of paramagnetic ions extracted with P; use of appropriate delay times and the minimum number of scans; and measurement of T1 values whenever possible.