Geogrids can be identified as comparatively low cost alternative for pavement construction on soft subgrades. However, comprehensive mathematical models are yet to be developed to define the composite behaviour of geogrid pavements. Therefore, it is required to perform pavement model testing to understand the performance of geogrid reinforcements under different pavement and geogrid conditions. Hence, this research study aims to compare the behaviour of composite geogrid in a weaker subgrade (CBR <3%) by performing two pavement model tests under repeated loading conditions. A steel test box with length, width and height of 1m, 1m, 1.2m respectively were used to construct both models with a subgrade having 500mm thickness and 3% CBR. A 200mm thick granular layer was compacted on top of the subgrade by achieving 91% degree of compaction. One model was selected as the control section and a composite geogrid at the base subbase interface was included in the other one. Both pavement models were tested for more than 100,000 repeated load cycles using a 200mm diameter plate on top of the granular base to simulate a tyre pressure of 550kPa. The results demonstrated that the inclusion of a composite geogrid significantly reduced the rutting depth of the granular layer and achieved a Traffic Benefit Ratio (TBR) of 5 at 50mm rutting. Furthermore, a significant reduction of pressure transmission to the subgrade by the composite geogrid was observed.