Lanthanide shift reagents have been used extensively in multinuclear magnetic resonance (NMR) applications in order to obtain information regarding ion distribution and transport in cellular systems. The aqueous reagents used in this study were Dy(PPP)27−, Tm(PPP)27−, Dy(TTHA)3−, Dy(PcPcP)27−, and Dy(DOTP)5−, where Dy3+and Tm3+represent dysprosium and thulium ions and PPP5−, TTHA6−, PcPcP5−, and DOTP8−denote the triphosphate, triethylenetetraminehexaacetate, bis(dihydroxyphosphinyl-methyl)phosphinate, and l,4,7,10-tetrazacyclododecane-N,N′,N″,N‴-tetrakis(methanephosphonate) ligands, respectively. The apparent size and shape of Li+-free RBCs (red blood cells), studied by both scanning electron microscopy and Coulter counter methods, were unchanged by the presence of the above shift reagents at concentrations lower than 10 mM. However, Li+incubation changed both the shape and size of RBCs. The rates of Na+-Li+exchange in Li+-loaded RBCs measured by 7Li NMR spectroscopy in the presence of Dy(PPP)27−, Tm(PPP)27−, or Dy(PcPcP)27−were significantly higher than the rates measured in the absence of shift reagents by atomic absorption or in the presence of Dy(TTHA)3−or Dy(DOTP)5−by 7Li NMR spectroscopy. 31P and19F NMR measurements of the membrane potential of Li+-free RBCs revealed that the shift reagents studied (except for Dy(TTHA)3−) do change the membrane potential, with the most negatively charged reagents having the largest effect. Thus, shift reagents must be used with caution in physiological NMR studies and in particular RBC applications. © 1990, American Chemical Society. All rights reserved.