Continental intraplate volcanic systems, with their locations far from plate tectonic boundaries, are not well understood: the crustal and lithospheric mantle structure of these systems remain enigmatic and there is no consensus on the mechanisms that cause melt generation and ascent. The Cenozoic saw the development of numerous volcanic provinces on the African plate, including within the Central Hoggar, located in Northwest Africa, part of the Tuareg shield. The magmatic activity began at approximately 34 Ma and continued throughout the Quaternary. In order to understand the origins and potential mechanisms that generated the intraplate volcanic activity in the Central Hoggar we aim to image the subsurface architecture, in terms of electrical resistivity, from the surface to the lithospheric mantle. To do so we use magnetotelluric measurements from 40 locations to generate a 3-D electrical resistivity model, over an area of about 100 km by 160 km. Low-resistivity features (i.e., conductors) are observed in the crust that are narrow, linear structures congruent with the boundaries of terranes and prominent fault zones (e.g., Azrou N’Fad). They likely reflect the Pan-African mega-shear zones, which were reactivated throughout the tectonic evolution of the region. The model reveals that these faults are lithospheric-scale. The low-resistivity features likely represent the signatures of past fluid pathways and mineralization. A deeper low-resistivity feature is observed in the upper lithospheric mantle directly beneath the Manzaz and Atakor volcanic districts. It may represent local, small-scale metasomatism of the sub-continental lithospheric mantle, and low-percent partial melting, that sits above a regional, large-scale asthenospheric upwelling associated with the Hoggar swell. It is likely the origin point of the fluids responsible for the overlying anomalies. The results highlight the control of the lithospheric-scale, mega-shear zones on the spatial distribution of the recent Cenozoic volcanic activity, which was influenced by the location of pre-existing structural weaknesses.