Based on the theories of Portis and of Castner 50 years ago, different continuous-wave measurement procedures for analyzing the microwave saturation power dependence of inhomogeneously broadened EPR lines were developed. Although these procedures have been refined, they still use only a few selected points on the saturation curve. A non-linear least-squares procedure for analyzing the microwave-power dependence of inhomogeneously broadened lines using all data points on a saturation curve has been developed. This procedure provides a simple alternative method to obtain magnetic relaxation data when the more direct pulse-saturation techniques are not available or are less suitable. The latter includes applications of quantitative EPR such as dosimetry. Then microwave saturation data should be obtained under conditions similar to those used in the quantitative measurements, which are usually made on first derivative spectra recorded using continuous-wave spectrometers. Selected applications to benchmark literature data and within the field of EPR dosimetry are discussed. The results obtained illustrate that relaxation times comparable to those yielded by various pulse-saturation EPR techniques can be obtained. It appears as a systematic feature that, whenever the pulse EPR data are fitted using bi-exponential functions, the shortest relaxation times obtained are those that correspond best to those measured using the current continuous-wave saturation method.