Self-sensing property of concrete is mostly assessed using small specimens without reinforcement, which may be misleading for real-time structures. To better simulate the actual field conditions, this study examined self-sensing of damage in large-scale reinforced concrete beams tested under four-point bending. During flexural testing, special attention was paid to the self-sensing capability of shear failure, since this type of failure occurs suddenly and catastrophically. Inadequate shear reinforcements were used to increase shear failure possibility of 100 x 150 x 1000 mm(3) (width x height x length) reinforced large-scale beams and beams were produced with a high shear span (a= 350 mm) to effective depth (d = 125 mm) ratio of 2.8. To increase the electrical conductivity of large-scale beams, chopped carbon fibers (CF) and multi-walled carbon nano tubes (CNT) were used. Instantaneous self-sensing recordings were made using brass electrodes embedded in different shear spans of beams. In addition to conducting self-sensing evaluations, researchers also investigated the effects of CF and CNT particles on the mechanical properties/structural behavior of large-scale beam specimens with the proposed reinforcement configuration. Results showed that compared to CNT, CF usage significantly improved the load carrying capacity and ductility, resulting in bending mode of failure even with indaquate shear reinforcement. Shear damage was successfully self-sensed in all tested beams, although all CF-based specimens started self-sensing from the beginning of loading with significantly higher changes in electrical resistivity results, unlike CNT-based specimens. Conductive network of CF-based specimens seemed to be disturbed more easily at high load levels. CF usage seems like a better option compared to CNT given its lower cost and easier mixability. (C) 2018 Elsevier Ltd. All rights reserved.