Deoxygenation of erythrocytes produced marked changes in their 31P nuclear magnetic resonance spectra in the superconducting spectrometer. Most significantly, all intracellular and extracellular phosphates underwent downfield shifts. In fully deoxygenated blood the extracellular phosphates showed downfield shifts that were dependent upon packed cell volume, when added pyrophosphate was used as a measure of extracellular chemical shift behavior. This effect on extracellular signals was attributed to the paramagnetic contribution of deoxyhemoglobin to the "bulk" magnetic susceptibility of the red cell suspension. Line broadening was observed in deoxygenated whole cell systems but not in hemolysates, as a result of paramagnetic susceptibility gradients across the cell membrane. The degree of downfield shift upon deoxygenation was of different magnitude for each intracellular phosphate [2-P of 2,3-diphosphoglycerate (2,3-DPG) greater than 3-P of 2,3-DPG greater than inorganic phosphate greater than ATP phosphates], independent of packed cell volume but dependent on the degree of deoxygenation of hemoglobin. When deoxygenation shift effects in adult cells were compared to those of cord blood cells containing 70% fetal hemoglobin, it was found that 45% of the 2,3-DPG shift effects were attributable to binding of that compound to hemoglobin. By use of a nonphysiologic phosphate analogue, methylphosphonate, as an internal reference, it was found that an increase in pH of deoxy cells contributed to the downfield shift of inorganic phosphate. In hemolysates, the methylphosphonate - inorganic phosphate chemical shift difference was found to be pH dependent, with a sensitivity of (-) 0.39 pH unit/ppm, independent of the hemoglobin oxygenation state.