The use of EGS, a general purpose Monte Carlo electron-photon transport system, has been investigated for low Z materials with incident electron energies as low as 50 keV while tracking electrons down to energies as low as 1 keV. It was found that: (1) the default electron step-size algorithm had to be modified, (2) care had to be taken to avoid turning off multiple scattering by using too short step-sizes; and (3) suitable cutoff energies for secondary electron production had to be used. Under these circumstances, it was found that for electron detector response functions and for electron transmission and reflection coefficients for slabs of material, EGS produces results which are in reasonable agreement with the calculated results of ETRAN and with experimental data. In the low Z cases investigated, no differences attributable to the different multiple scattering formalisms used in the two codes could be found although, as expected. EGS does predict too much bremsstrahlung production for electrons below 2 MeV kinetic energy because it ignores the Elwert correction factor. The straggling in the energy-loss distribution predicted by ETRAN using the Landau formalism was found to be in good agreement with the energy distribution calculated by EGS which includes energy-loss straggling by explicitly taking into account energy loss from random discrete events. A minor logic error was found in EGS when production of low energy secondary electrons was considered (less than 80 keV for water), although the observable effects were found to be small. Coding for a short routine to provide electron step-sizes of equal energy loss is presented.