An increasing interest in both efficient electric machines and more extensive control strategies demands evermore faster simulation tools. In that context, Fourier-based models, which combine low computational times with a high accuracy, have already proven their value. However, even Fourier-based models may encounter problems related to CPU usage. To cope with these problems the authors present a number of simple techniques to reduce the computational effort of Fourier-based models for synchronous machines. The techniques are based on simplifying the studied geometry and a qualitative analysis of the machine's time-and spatial-harmonic content. The proposed techniques are validated and a benchmark test has shown that a great reduction in computational burden can be achieved without significant loss of accuracy. The preliminary harmonic analysis gives, by far, the largest reduction of the computational burden.