By the principle of quasi-static calibration of shock wave pressure sensor, the comparison calibration accuracy depended on the accuracy of fitting model between the height of drop-hammer and pressure. The paper applied the RBF neural network to set up the neural network model whose input was the height of drop-hammer and output was the shock wave pressure peak. Choosing a certain type of standard pressure sensor, the calibration experiment was carried out in the range of 7MPa-30MPa. By analyzing the test samples, the result indicated that the maximum relative error of the neural network model is not more than 0.04% and a higher magnitude than the polynomial fitting model and exponential fitting model. The uncertainty introduced by the fitting model between the height of drop-hammer and pressure is an important component of dynamic measurement uncertainty of shock wave pressure sensor. Building the neural network fitting model between the height of drop-hammer and shock wave pressure peak will lay the foundation of improving the measurement accuracy of shock wave pressure sensor.