Abstract : The Zagros mountain belt situated on the northern margin of the Arabian plate, is one of the youngest belts of continental collision. This belt was built by the collision of the Arabian plate with the central Iranian micro-continent. A seismological experiment, called “ Zagros 2000-2001 ”, was realized by collaboration between LGIT and IIEES to study the lithospheric structure beneath this collision belt and some part of the central Iranian block. We used the data obtained during this experiment to characterise the structure of the crust and the lithospheric mantle under the stations network. The analysis of receiver functions was used to investigate the variations of the crustal thickness under the network. By this analysis, we have found a thickening by 20 km over a region of about 100 km width just on the northeast of the MZT (“ Main Zagros Thrust “). The average crustal thickness of 45 and 40 km were found respectively for the Zagros and central Iran. We've then proposed a crustal model in which the Moho was constrained using the results obtained by the receiver functions analysis. This model was tried to be compatible with gravimetric data. The thickening of the crust on NE of the MZT was interpreted to be related to overthrusting of the central Iranian crust onto the Zagros crust. We have also characterised the upper mantle structure to depth of 350 km by inversion of more than 5000 arrival times of teleseismic P waves. The results of this inversion show a fast upper mantle beneath the Zagros and a slow one beneath central Iran. We may relate the presence of a slow upper mantle beneath central Iran to delamination of the lithospheric mantle. In excess, the presence of a slow and light mantle beneath central Iran may explain the relatively high elevation of the Iranian plateau. The analysis of splitting of teleseismic S-waves show that the lithosphere of the Zagros and central Iran are different in terms of seismic anisotropy. This analysis underlined the absence of splitting beneath the Zagros as opposed to some regions in central Iran where we've observed an average splitting of more than 1 sec. On the other hand, we didn't observe any clear relation between the direction of the fast axes of the observed splitting and the direction of relative or absolute plate motions. The observed splitting should be produced by some “ fossil anisotropy ” into the lithosphere of central Iran. This anisotropy may be related to a tectonic episode before the continental collision between the two plates.