Geogrid reinforcement is widely used to improve flexible pavements over weak subgrades, yet full-scale evidence on the effect of reinforcement location under repeated wheel loading remains limited. This study investigated the influence of geogrid placement on pavement response using full-scale Accelerated Pavement Testing. Three pavement sections were constructed over a weak subgrade with a California Bearing Ratio of approximately 5%: an unreinforced control section (A1), a section reinforced at the subgrade-subbase interface (A2), and a section reinforced at the top of the granular subbase layer beneath the asphalt binder course (A3). All sections had identical layer configurations and were subjected to 3,000,000 load repetitions under field temperature conditions. Performance was evaluated using rut depth, transverse surface deformation profiles, vertical stress near the top of the subgrade, and tensile strain at the bottom of the asphalt layer. Reinforcement improved all measured responses, with location strongly affecting performance. Section A3 showed the best behavior, reducing rut depth from 21.5 mm in A1 to 12.1 mm, while giving the lowest final vertical stress, 46.73 kPa, and tensile strain, 154.7 με. These results indicate that placing geogrid at the top of the subbase layer was most effective under the tested full-scale loading conditions.