The Earth's magnetic field is generated through magnetohydrodynamic processes within the Earth's fluid outer core. Several factors that may influence these processes, such as variations in heat flow across the core-mantle boundary, are believed to have varied through geologic time, triggering speculation that the Earth's magnetic field may have been significantly different in the geological past. Based on painstakingly difficult paleo-intensity experiments, it has, for example, been suggested that the Earth's dipole moment was much weaker during the Mesozoic, and perhaps related to different mantle convection. The variable quality of published paleo-intensity estimates makes global compilations difficult, and extreme care must be taken not to interpret problematic data in terms of geodynamic features. In fact, recently available reliable paleo-intensity data suggest that the paleo-strength during early-middle Cretaceous may have been comparable or even higher than the present intensity, and not anomalously low as suggested in previous studies. The identification of any long-term features of virtual dipole moment (VDM) seems to rely heavily on the criteria for accepting paleo-intensity data, which casts serious doubts on the quality and geomagnetic significance of the majority of existing data. In a recent analysis, Thomas and Biggin <2003>-based on so-called self-consistency selection criteria-concluded that the relatively low intensity during the Mesozoic is a real phenomenon.