Biological scanning electron microscopy is increasingly performed at low beam energies in order to improve image contrast, reduce charging artifacts, and minimize beam induced damage to the sample. It is natural then to wish to also use these same low accelerating voltage for X-ray microanalysis using an energy dispersive spectrometer (EDS) but a variety of fundamental physical effects affect the performance that can be achieved. An X-ray photon can only be emitted when the incident beam Eo energy exceeds the critical excitation energy Ecrit for that line. As the beam energy is reduced the number of elements that can be excited falls and it is becomes necessary to use L- and M-lines rather than the K-lines accessible at higher energies. At 5keV, the upper limit of ‘low voltage’ microscopy, K-lines can be excited from elements up to calcium, L-lines can be detected up to cesium, and the rest of the periodic table is available using M-lines. The entire periodic table is therefore, in principle, available at an overvoltage U>2, where U = E0/Ecrit But at lower energies the number of accessible excitations falls and some elements cannot, with current technology, then be analyzed and for general purposes an incident energy lower than about 3keV is probably too limiting to be useful.