Using split Hopkinson pressure bar (SHPB) to carry out the Brazilian disk test, for the calculation of the splitting tensile strain rate of the specimen, the method in the existing literature only considers the influence of the tensile stress, while in the actual splitting tensile process, the combined effect of compressive stress and tensile stress leads to a large error between the calculated strain rate and the measured strain rate. Therefore, the calculation of splitting tensile strain rate by the existing method is not accurate. To solve this problem, a new analytical method considering the tension–compression coupling was proposed in this study. The true strain rate of specimens can be obtained using the classical three-wave method of SHPB test and the basic performance parameters of bars and specimens. The results of this method were verified by numerical calculation and experiment. The results show that the strain–time history curve calculated by this method is highly consistent with the numerical results in terms of numerical verification, and the error range of strain rate is 1.2–8.6%. However, the error range of strain rate obtained using the existing method is 34.4–41.2%. The rising stage of strain–time history curve calculated by this method is highly consistent with the experimental results in terms of experimental verification, and the error range of strain rate is 1.7–14.7%. While the error range of strain rate obtained using the existing method is 47.2–58.5%. The existing method of splitting tensile strain rate will overestimate the dynamic tensile strength of material and make engineering design in a dangerous state.