Abstract: Neutron kinetic parameters such as effective delayed neutron fraction βeff and mean generation time of prompt neutron are the bridges that between reactivity (in unit of dollar) measured by experiment and reactivity by calculation. Accurate measurement and calculation of subcritical reactivity is an important problem in the development of accelerator driven subcritical reactor and control rod calibration in subcritical condition. To compare subcritical reactivities between experiment and calculation, experimental measurement methods that based on fundamental prompt neutron attenuation constant α need neutron kinetic parameters from calculation using fundamental prompt neutron flux density of α mode. In this paper, based on open source code OpenMC and the kα iteration method, taking the average time absorption weight of the neutron track length as a parameter of the kα iteration factor, and neglecting delayed neutron in the transportation process, the OpenMCPA module with the function of calculating prompt α eigenvalue problem was developed. The ability of the code to calculate prompt α eigenvalue problem was verified by using the Godiva derivative benchmark and the MUSE4 subcritical experiment device. The results indicate that for the Godiva derivative benchmark, results of OpenMCPA are in good agreement with those of TART code. In addition, under the same computing hardware, the calculation speed of OpenMCPA is faster than that of MCNP4C, and the subcriticality that can be calculated is deeper. For the MUSE4 subcritical experiment device with complex geometry and materials, the relative error of prompt α eigenvalue between the calculated value of OpenMCPA and the experiment is less than 0.5%, and the absolute error between the keff by regression calculation using calculated prompt α eigenvalue and the keff from experiment is less than 0.02%. The developed OpenMCPA can meet the requirements of prompt α eigenvalue and kinetic parameters calculation for subcritical systems.