The central Andes is the highest and widest region of Andean Cordillera in South America. This region currently forms a high plateau with an average elevation of 4000 m. However, the cause of central Andean deformation remains unclear as this area has not experienced any collision or accretion events in the last ~200 Ma. In addition, it is puzzling that significant crustal shortening occurred in this area but not in regions to the south. In this study, 2D thermal-mechanical models are used to study continental deformation above an active subduction zone. The regional-scale models extend to a depth of 1200 km, and include a continent that moves toward the trench, consistent with the Cenozoic tectonic setting of the Andes. The models demonstrate that folding of the oceanic slab as it encounters the high-viscosity lower mantle and trenchward continental motion provide the primary controls on crustal shortening and plateau growth. Folding and buckling of slab in the lower mantle cause the slab dip to decrease at shallower depths, enabling compression of the continent. This is enhanced by a faster overriding plate velocity. Furthermore, a weaker continental lithosphere rheology or stronger interface friction at the plate boundary increase the amount of shortening. Our models are in agreement with the first order observations of crustal shortening, crustal thickness and surface uplift in the central Andes. In addition, we infer that a weaker interface and/or stronger continent in the southern Andes may explain the lack of significant deformation there, despite the similar tectonic setting.