In this study, a series of numerical pullout tests was performed using the discrete-element method (DEM) to investigate the micromechanical behavior of geogrid pullout and the effect of geogrid tensile stiffness. Geogrid-soil interaction during pullout was investigated not only through displacement fields and force chains inside the soil but also via the quantitative displacement and force distributions along the geogrid. The active and inactive zones were defined based on the displacement fields in the soil, and the displacement corresponding to the boundary between the active and inactive zones was found to be 1.25 mm in this study. The higher the geogrid stiffness, the larger was the thickness and length of the active zone mobilized in the soil under the same pullout displacements. The thickness and length for the geogrid with largest tensile stiffness were 150 and 500 mm (full length of geogrid), respectively. Reorientations of contacts and forces within both the inactive and active zones were further visualized based on the Fourier-series approximation. The part of the geogrid that experienced displacement of more than 1.25 mm (used for the definition of the active zone) is defined as the affected part, and its corresponding length is defined as the affected length. The stiffer geogrid was found to activate the affected length into the full range more rapidly than the less stiff one prior to failure. The tensile forces at the load ends of stiff geogrids were larger than those at the load ends of extensible geogrids at all given pullout displacements.