Differences in magnetic susceptibility between various compartments in heterogeneous samples can introduce unanticipated complications to NMR spectra. On the other hand, an understanding of these effects at the level of the underlying physical principles has led to the development of several experimental techniques that provide data on cellular function that are unique to NMR spectroscopy. To illustrate some key features of susceptibility effects we present, among a more general overview, results obtained with red blood cells and a recently described model system involving diethyl phthalate in water. This substance forms a relatively stable emulsion in water and yet it has a significant solubility of 5 mmol/L at room temperature; thus, the NMR spectrum has twice as many resonances as would be expected for a simple solution. What determines the relative intensities of the two families of peaks and can their frequencies be manipulated experimentally in a predictable way? The theory used to interpret the NMR spectra from the model system and cells was first developed in the context of electrostatics nearly a century ago, and yet some of its underlying assumptions now warrant closer scrutiny. While this insight is used in a practical way in this article, the accompanying article deals with the mathematics and physics behind this new analysis.