Abstract: In order to observe the effected strip of geogrid-reinforcement, a series of geogrid pull-out tests are conducted in a test box which has one Plexiglas wall, applied 4 different vertical pressures on each specimen prepared respectively with 6 granular soils, with each having a different particle gradation. Through the Plexiglas wall, the displacements of the soil particles are observed with the help of the embedded marks which are the pinpoints wrapped by rubber wire coat. The displacements are found of the soil particles in the region from the top surface of the geogrid up to a certain height in the soil mass. This region is herein referred to as the reinforcement-effected strip. It is demonstrated that δ, the depth of the geogrid-effected strip, is related to the particle gradation rather than to the vertical pressure loaded on the specimen while d₅₀, the average diameter of soil particles, is greater than 0.83 mm. As d₅₀ gets larger, δ increases rapidly when d₅₀ is less than 1.05 mm and increases slowly when d₅₀ is greater than 1.05 mm. Especially, while d₅₀ is greater than 1.65 mm, δ has a good linear increasing relationship with d₅₀. Based on the test results, a method considering the reinforcement-effected strip, RESM for short here, is suggested to analyze the stability of a reinforced soil slope, in which the function of a geogrid reinforcement is assumed just only to increase the cohesion of the soil within the reinforcement-effected strip. Therefore, a reinforced soil slope is regarded as a layered soil slope in RESM, leading to a more simple stability analysis of a reinforced soil slope. The safety factors calculated by RESM are approximately equal to those by the strength reduction method while the reinforcement spacing is not greater than 0.6 m and the geogrid tensile strength is greater than 20 kN/m.