Fibers Reinforced Polymer (FRP) bars are being increasingly used recently as a replacement of conventional steel reinforcement in structures subjected to harsh environmental conditions due to their non-corrodible characteristics. However, little attention has been given to deep beams reinforced with longitudinal FRP bars and without web reinforcement. Usually, the Strut and Tie Model (STM) approach is used for the design and analysis of deep beams reinforced with steel reinforcement as recommended by many codes of practice; however, it is not clear whether those models are applicable to deep beams reinforced with FRP bars. Therefore, this study investigates the accuracy of the STMs adopted by ACI 318-14, Eurocode 2 and CSA S806-12 codes to assess the strength of deep beams reinforced with longitudinal FRP bars. For this purpose, a total of 52 FRP reinforced deep beams was collected from previous experimental investigations, and their failure loads were predicted using the aforementioned STMs, and compared with the actual failure load obtained from experiments. In general, the results have shown that both ACI 318-14 and Eurocode 2 STMs can result in unsafe design, since they overestimate the shear strength of FRP reinforced deep beams considerably. While the STM proposed by CSA S806-12 can result in uneconomic design, since it underestimates the shear strength considerably. In order to determine the appropriate efficiency factor to be used for the concrete strut, a regression analysis was conducted using the data base. Based on which, a new efficiency factor was proposed and incorporated into the Strut and Tie Model. The new STM was then tasted using the database, and it proven to be much more accurate than the other STMs.