In this paper the K-stiŠness Method as originally proposed by Allen et al. (2003) is re-examined using a total of six new case studies-ˆve from Japan and one from the USA. A common feature of the walls in this new data set is that the walls were all constructed with a vertical face and a granular backˆll. However, the walls varied widely with respect to facing type. This new data set together with data for vertical walls previously published by Allen and Bathurst (2002a,b) and Allen et al. (2002) is now used to isolate the eŠect of the facing stiŠness factor on reinforcement loads and to adjust the original equation that was developed to calculate its value. The paper also shows that predicted reinforcement loads using the current AASHTO Simpliˆed Method in the USA and the current PWRC method in Japan give the same reinforcement load predictions, and both grossly over-estimate the values deduced from measured strains. The new data set is used to slightly reˆne the estimate of the facing stiŠness factor used in the original K-stiŠness Method. The original and modiˆed K-stiŠness Method are demonstrated to quantitatively improve the estimate of the magnitude and distribution of reinforcement loads for internal stability design of vertical-faced geosynthetic reinforced soils walls with granular backˆlls when compared to the current American and Japanese methods.