This study compared the performance of three aggregate layerings, commonly used in the construction of unbound forest roads in Ireland, when they were subjected to repeated loading in a new large-scale test rig. These layerings comprised (i) a layer of uncrushed, granitic, sandy gravel - a good quality road aggregate (ii) a layer of shale - a poor quality aggregate, and (iii) a layer of crushed limestone an excellent quality aggregate with a wet mix macadam (WMM) grading on top of a poor quality shale sub-base layer . The repeated load testing rig was designed and constructed to test different surface or completion layering thicknesses of the aggregates over a common formation or subgrade material of silty sandy soil. This testing was achieved by surface loading the aggregates through a 200 mm-diameter rubber pad - attached to a hydraulic actuator on the test rig - for up to 150,000 load applications. The subgrade pressures and surface deflections were measured at applied stresses of 500 kPa, 750 kPa and 1000 kPa. The good quality granitic aggregate performed much better than the poor quality shale aggregate under the repeated loading and is suitable as a completion material for use in unbound forest roads. The shale aggregate can be used in unbound forest roads as a sub base material.