Abstarct: The magnetotelluric method is an electromagnetic geophysical method with high penetration depth, which determines the distribution of subsurface electrical resistivity with respect to depth by using the natural electromagnetic fields of the earth and outside the atmosphere that enter the earth. The high penetration depth of the magnetotelluric method has led to its use in the exploration of deep structures, such as the exploration of geothermal and hydrocarbon resources In this study, the magnetotelluric data of Busheli region of Ardabil province have been analyzed with the aim of exploring geothermal resources in this region. The data of 45 magnetotelluric measurement stations in this area have been analyzed. First, by drawing the curves of specific resistance and the phase data of each station, the data containing the noise has been removed. Then, using dimensional analysis methods such as skewness, ellipticity, and impedance tensor polar diagrams, the dimensional analysis of the data has been done. The results of these analyzes show that in most stations, subsurface geoelectrical structures on a regional scale are two-dimensional in high to medium frequencies (about 200-1 Hz) and even sometimes one-dimensional and in medium to low frequencies. (1-0.01) are three-dimensional. In fact, towards greater depths, geoelectrical structures are more complex and have a three-dimensional nature. Considering the approximate extension of geoelectrical structures, some two-dimensional profiles in the design area and two-dimensional inverse modeling of the data for the purpose of two-dimensional interpretation have been done along these profiles. The results of the two-dimensional inversion using the distribution of the real specific resistance of the subsurface have separated the geological structures related to the geothermal resources and relatively determined the location of the alteration zones and the geothermal source. In the next step, to better study the geothermal source, the three-dimensional inversion operation has been implemented on the major part of the data, taking into account the position of the stations in a suitable network. Examining the 3D inversion results relatively better than the 2D results show the location of the alteration zones and the corresponding geothermal source, and considering the geological information, these structures can be well interpreted.