Following the paper by Fraser-Smith et al. (1990), many scientists have focused their research on the ULF geomagnetic field pulsations in the hope of finding possible anomalous signals caused by the seismic activity. Thereafter, many papers have reported ULF geomagnetic field polarization ratio increases which have been claimed to be related to the occurrence of moderate and strong earthquakes. Even if there is no firm evidence of correlation between the polarization ratio increase and seismic events, these publications maintain that these "anomalous" increases are without doubt precursors of pending earthquakes. Furthermore, several researchers suggest that these seismogenic signals may be considered a promising approach towards the possibility of developing short-term earthquake prediction capabilities based on electromagnetic precursory signatures. On the contrary, a part of the scientific community emphasizes the lack of validation of claimed seismogenic anomalies and doubt their association with the seismic activity. Since earthquake prediction is a very important topic of social importance, the authenticity of earthquake precursors needs to be carefully checked. The aim of this paper is to investigate the reliability of the ULF magnetic polarization ratio changes as an earthquakes' precursor. Several polarization ratio increases of the geomagnetic field, which previous researchers have claimed to have a seismogenic origin, are put into question by a qualitative investigation. The analysis takes into 2 account both the temporal evolution of the geomagnetic field polarization ratio reported in previous papers, and the global geomagnetic activity behaviour. Running averages of the geomagnetic index K p are plotted onto the original figures from previous publications. Moreover, further quantitative analyses are also reported. Here, nine cases are investigated which include seventeen earthquakes. In seven cases it is shown that the suggested association between the geomagnetic field polarization ratio increases and the earthquake preparation process seems to be rather doubtful. More precisely, the claimed seismogenic polarization ratio increases are actually closely related to decreases in the geomagnetic activity level. Furthermore, the last two investigated cases seem to be doubtful as well, although a close correspondence between polarization ratio and geomagnetic activity cannot be unambiguously demonstrated.