Experimental studies were carried out to obtain the load-settlement response of a model square footing resting on unreinforced and reinforced granular beds. The response was obtained for two cases: (a) geogrid-reinforced sand layer, and (b) geogrid-reinforced layered system consisting of aggregate layer overlying a sand layer. The parameters considered in the experimental study include the thickness of the aggregate layer, the depth of geogrid reinforcement placed in sand layer and in aggregate layer, width of the reinforcement, and relative density of bed. Plate vibrator was used to compact uniform sand beds to relative densities equal to 50 % and 70 % inside large-size test chamber of dimensions equal to 1 m × 1 m × 1 m (in length, in width, and in depth). Load was applied on square footing using a 100 kN capacity actuator in displacement-controlled mode, and the improvement in the load carrying capacity of the footing resting on reinforced sand layer and layered system was quantified in terms of load improvement factors. In addition, the optimum embedment depth and width of reinforcements were proposed for various cases considered in the study. The optimum depth of reinforcement for the case of aggregate layer overlying sand layer decreased to 0.30 times the width of the footing from 0.45 times the width of the footing for sand only case. © 2016, Springer International Publishing Switzerland.