Glass or carbon fibre-reinforced polymer (FRP) structures have attained a relevant role in modern production thanks to their lightweight, high resistance to corrosion and fatigue, and relative ease of manufacturing complex shapes. Nevertheless, several types of defects, such as voids, matrix fractures, delaminations and impact damages, which might significantly reduce the useful life of composite-made structural components, are often difficult to be identified. Conventional non-destructive methods, such as ultrasonic testing, are routinely used to monitor composite constructions with good results. However, an inspection of large structures requires alternative and faster NDT methods. Among these, air-coupled ultrasonic testing (ACUT) and active infrared thermography (AIRT) are generally considered the most promising ones. In this paper, ACUT and AIRT have been compared by performing experimental tests on two reference composite panels with synthetic defects. Polytetrafluoroethylene (PTFE) and aluminium inserts with different sizes and located at various depths have been used as artificial defects. The relative merits and drawbacks of the two techniques have been highlighted in terms of defect size sensitivity, maximum depth that can be detected, and thermo-physical properties of the reflector. A qualitative and quantitative analysis has been carried out to evaluate the ability of the two techniques in retrieving the actual size of the subsurface defects.