An indirect tensile testing method is proposed to measure the full dynamic tensile stress-strain curve of low strength brittle solids. In this method, the flattened-Brazilian disc (FBD) sample is loaded by modified split Hopkinson pressure bars (SHPB) system. Low amplitude dynamic forces were measured with a pair of piezoelectric force transducers embedded in the incident bar and the transmitted bar. The evolution of tensile stress at the center of the disc sample was determined through finite element analyses using the measured stress in SHPB as inputs. In a traditional Brazilian test, a strain gauge is mounted at the center of the specimen to measure the tensile strain, which is difficult to apply for low strength brittle materials. Thus, two types of non-contact methods, the Digital Image Correlation (DIC) technique and the Laser Gap Gauge (LGG), were used to measure the strain. The DIC method was used to monitor the displacement and the strain map of the specimen during the test, from which the strain at the center of the specimen can be obtained. The accuracy of the DIC results was assessed, and the displacement and strain uncertainties of our system were 0.003 mm and 0.003, respectively. LGG was used to monitor the expansion of the disc perpendicular to the loading axis, from which the average tensile strain is deduced. The numerical simulation revealed that the tensile strain at the center of the specimen is proportional to the average tensile strain and that the ratio is not sensitive to the material elastic parameters. The strain measured through LGG was compared with that measured by the DIC method using photos captured with a synchronized high-speed camera. The result of the LGG method was 20 % smaller than that of the DIC process. However, the latter was limited by the number of frames of the high-speed camera. The feasibility of this methodology was demonstrated using a polymer-bonded explosive (PBX).