There are several kinds of coupling mechanisms which can convert mechanical, chemical or thermal energies due to seismic or volcanic activities into electromagnetic energies. As a result of concentrated efforts in laboratory and theoretical research, the basic relationship between the intensity of electromagnetic sources and changes in mechanical, chemical and thermal state is becoming established. Also with the progress of the electromagnetic simulation techniques, it has been possible to evaluate in situ sensitivity. Based on this progress and also due to extensive improvement in measuring techniques, many field experiments have been performed to elucidate subsurface geophysical processes underlying the preparation stage, onset, and subsequent healing stage of earthquakes and volcanic eruptions. In volcanic studies, many studies have reported the measurement of electromagnetic signals which were successfully interpreted in terms of various driving mechanisms. Although there have been numerous reports about the existence of precursory electromagnetic signals in seismic studies, only a few of them could be successfully explained by the proposed mechanisms, whereas coseismic phenomena are often consistent with those mechanisms including the absence of detectable signals. In many cases, one or two orders of higher sensitivity were required, especially for precursory signals. Generally, electromagnetic methods are more sensitive to near-surface phenomena. It will be necessary to discriminate electromagnetic signals due to these near-surface sources, which often possess no relationship with the crustal activities. Further efforts to enhance in situ sensitivity through improvements in observation techniques and in data processing techniques are recommended. At the same time, multi-disciplinary confirmation against the validity of electromagnetic phenomena will inevitably be necessary. A Network-MT observation technique has been developed to determine large-scale deep electrical conductivity structure. In the method, a telephone line network or purpose-built long baseline cables are utilized to measure voltage differences with long electrode separations. Because of the averaging effect of the electric fields, static shift problems due to small-scale, nearsurface lateral heterogeneities can be alleviated. Several field experiments revealed regional scale deep electrical conductivity structures related to slab subduction or its M. Uyeshima (stagnation, which enable us to elucidate underlying physical processes caused by the slab motion. The technique can also be applied to monitor the electric potential field related to crustal activities. The annual variation of the potential field and electrical conductivity in the French Alps were interpreted to be caused by the annual variation of lake water level. The method was also used to monitor the regional scale spatio-temporal variation of the SP field and electrical conductivity before and at the onset of earthquakes and volcanic eruptions.