Abstract: The performance of reinforced concrete structures strengthened with fiber reinforced polymer (FRP) composites depends on the bond between FRP and concrete. Moisture, as one of the most common environmental factors, is very important for durability of bond between FRP and concrete. Therefore, the methods to simulate and predict the bond durability in moist environments are critical to the field application of FRP strengthening techniques. The Modified Double Cantilever Beam (MDCB) test was used to measure the interfacial fracture energy for the carbon fiber reinforced polymer (CFRP) plate debonding from concrete substrate under peel test (Mode I loading). A simple method was developed to directly measure the value of Interface Region Relative Humidity (IRRH). Using this relation and Virtual Crack Closure Technique (VCCT), the bond fracture energy was numerically determined as a function of IRRH for the FRP/concrete bond joints. Modeling the transportation and distribution of moisture in FRP strengthened concrete structure is particularly important because the interfacial adhesion between FRP and concrete can be susceptible to moisture attack. The moisture diffusivity and isotherm curve of each constitutive material (concrete, epoxy and FRP) were experimentally determined. The bond interface region relative humidity was obtained by the numerical diffusion analysis based on the measured material diffusion properties. A bond mechanism based deterioration model of bond interfacial fracture energy was proposed for FRP-concrete bond joints in moist environments. The bond interface region relative humidity was correlated to the bond fracture energy in this deterioration model. The IRRH-dependent interface separation-tractions were then derived in the frame of cohesive zone model (CZM). Such an IRRH-dependent interfacial separationtraction law was used to simulate the bond performance in moist environments. With the bi-linear separation-stress law in the cohesive zone model (CZM), the analytical solutions of interfacial stress, interface separation and ultimate load of the plated beam were obtained. A simplified explicit expression was also derived to determine the load-crack length relationship of peel test specimen in moist environments. The good agreement with experimental data indicates that the approach developed in this study is an efficient way to simulate bond durability in moist environments.