Currently, several different methods for designing geosynthetic reinforced soil walls are available in the literature. An evaluation of these methods can be carried on by a direct comparison with the observed response of instrumented walls in the field. This comparison is the main objective of this research work. Three case histories of geosynthetic reinforced soil wall, constructed in Brazil, were selected for this research. The monitored response of these structures registered the tension in each reinforcement layer during construction. The maximum values of reinforcement tension have been compared with the computed values from different design methods. Moreover, predicted tension values from numerical simulations were also compared to the measured values in each reinforcement layer in the instrumented field walls. These comparisons indicate that, in two of the three evaluated cases, the design methods based on limit equilibrium underestimated the maximum tension. This was noted to be particularly significant in the upper layers of reinforced walls compacted under high energy levels. The analytical method based on work conditions proposed by Ehrlich and Mitchell (1994) resulted in tension values higher than those registered in the field instrumentation, for the three selected cases. Numerical simulations predicted maximum tension in reinforcements with similar values than those from the field instrumentation. The Ehrlich and Mitchell (1994) formulation for predicting the vertical tension induced by compaction resulted coherent with computed values from numerical finite element method for the three walls evaluated herein.