||The MT method, like all resistivity methods that are based on measuring the electric field in the surface, suffer the so-called telluric or static shift problem manifesting itself in an unknown multiplier of the apparent resistivity (a constant shift on log-scale). This phenomenon is caused by resistivity in-homogeneities close to the electric dipoles. Severe topography can also lead to static shifts. Except for very high frequencies, the shifts are independent frequency. Static shifts can be extreme in geothermal areas in volcanic environments where resistivity variations are often huge. The problem is made even worse by the fact that the shifts are often not random. All soundings in large contiguous areas can be consistently shifted up or down. Extreme examples of this are presented here. It is sometimes claimed that the static shifts can be dealt with by resolving the shallow resistivity structure around the electric dipole by measuring at high enough frequencies. This would hold true if the earth can be considered as a pure Ohmic conductor. But at high frequencies other processes set in. Capacitance and induced polarization effects become important and lead to reduction of the electric field and consequently bias the apparent resistivity down at very high frequencies. Various techniques to use the MT data themselves to identify and correct for static shifts have been proposed and tried. These "pseudo" corrections are usually based on some spatial averaging and/or some statistical assumptions about the shifts e.g. that the shift multipliers are random and that the product of the shift multipliers of individual soundings is close to one for sufficiently many soundings covering large areas. It is shown here that this assumption is far from being true in geothermal areas in volcanic environments. Since 3D inversion of large MT datasets became available, some service vendors have claimed that the 3D inversion can cope with static shifts. By detailed modelling of topography, shifts of topographic origin can be modelled to some extent. But it has also been claimed that shifts due to shallow resistivity in-homogeneities can be dealt with by using fine model grids near the surface. The inversion would then introduce appropriate resistivity bodies at shallow depth. This has, however, not been demonstrated convincingly in the literature. The central-loop TEM method is only sensitive to the near surface resistivity structure and topography at very early times. At late times their effects have practically disappeared. Here it is shown that a joint inversion of MT and TEM data is a consistent and effective way to correct for static shifts in MT soundings and should be used in 1D and prior to 2D and 3D inversion. If static shifts are corrected for by TEM, topography should not be modelled in the inversion; that would account twice for shifts due to topography. It is the view of the author that, except in special cases like thick and homogeneous sediments close to the surface, MT data alone should be considered as incomplete data for geothermal exploration.