Fiber-reinforced polymer (FRP) systems that have a strong resistance against long-term deformation must provide improved serviceability to reinforced concrete (RC) members under sustained loads. Consequently, there is a need to develop a method for accurately predicting the time-dependent behavior of RC beams that are externally bonded with FRPs. However, there are very few previous studies that have been carried out or experimental results available, on the time-dependent behavior of RC beams externally bonded with FRP. In order to enable a reasonable prediction, correlations should first be clarified between the stress–strain relationship of the concrete, the reinforcement and the FRP that changes over time. By using these correlations, deflections under sustained loads should then be forecast. In this study, RC beams were fabricated for this purpose. Carbon reinforced polymer (CFRP) and glass reinforced polymer (GFRP) materials were bonded to the tension face of the two respective RC beams. The beams were then placed under sustained loads for 300 days. For the specimens that were externally bonded with FRPs and for the conventional specimen, the strain of the compression and tension reinforcement and the strain of FRP and deflection were measured respectively for comparison. In order to theoretically predict the time-dependent behavior of the RC Beam externally bonded with FRPs, creep coefficients for concrete and shrinkage strains were calculated by using the CEB-FIP and the ACI-209 Codes. For the method used to forecast the stress–strain relationships of the concrete, reinforcement and FRPs that change over time were theoretically clarified and were then compared with the experimental results. The deflection of the RC Beams externally bonded with FRP was predicted by using the ACI 318 Standard, EMM, AEMM, Branson’s method, and Mayer’s method. They were also compared to the experimental results. Subsequently, in the case of RC Beams externally bonded with FRPs under sustained loads, the proposed method proved that it is possible to accurately predict long-term deformations.