This paper investigates the calculation, modelling and interpretation of efficiency maps for electrical machines. The efficiency maps are calculated using a finite-element based mapping of losses, torque and flux-linkage as a function of the d- and q-axis currents and speed. For modelling efficiency maps, it is shown that a number of key loss types can be described in the form Tmωn. The effect of each of these losses on the shape of the efficiency map is then explored. It is found that practical efficiency maps can be approximated using a series of such terms which leads to a better understanding of the losses in the machine. The above results are validated using the loss and efficiency maps of three example machines.