In the past two decades, the use of Near-Surface Mounted (NSM) Fibre Reinforced Polymer (FRP) strengthening technique has increased in the retrofitting of damaged structures such as buildings, bridges, storages, etc. In the NSM technique, epoxy resins are usually used as a bonding material and have an important effect on the bond performance between the FRP and concrete. Exposure of these resins to increment of temperatures (i.e. around or beyond the glass transition temperature, T-g) may lead to changes in their mechanical properties, so that the performance and effectiveness of the strengthening system can be affected. In the present work, an experimental programme is performed to study the effect of temperature on the behaviour of NSM FRP strengthened RC beams. Unlike to most of the existing literature, where performance of the strengthening system under fire temperature was studied, the aim of this study is to consider a high ambient temperature (i.e. similar to temperature during the summer days). Six specimens have been tested up to failure in two different series. Each series includes a control beam and two NSM CFRP strengthened beams with two different areas of CFRP. In the first series, the specimens have been tested at room temperature, whilst those of second series have been tested at 40 degrees C. Results of the effect of temperature on the flexural response of strengthened RC beams with different NSM CFRP ratio are evaluated in terms of flexural performance, failure mode and bond performance along the FRP laminate. According to experimental results, the stiffness of specimens was not significantly affected by temperature. Besides, large laminate length avoided FRP debonding in strengthened specimens and CFRP rupture was the observed failure mode, irrespective of the testing temperature. Finally, bond resistance was assessed with available analytical procedure. Comparison between analytical predictions and experimental results confirm the procedure to be conservative, as no end debonding was neither predicted nor experimentally observed.