In this paper, the extraction of the relative permittivity from transmission line measurements by neglecting and considering the metal losses is performed. It is observed that the extracted permittivity obtained in both cases allows to achieve good model-experiment correlation for S-parameters. However, when performing time-domain simulations, if the contribution of the metal losses on the phase delay is neglected on the extraction of the permittivity, the simulated propagation of the signal presents an extra delay when compared with the measurements. Conversely, simulations are in agreement with voltage waveforms obtained from experimental data when appropriately considering the metal losses.