Abstract: The treatment planning system (TPS) of boron neutron capture therapy (BNCT) is a software composed of a series of modules. It is used to evaluate the dose distribution based on medical information and the characteristics of the therapeutic neutron source, assisting physicians in formulating precise radiation treatment plans. The TPS enables better tumor control and the protection of normal tissues, thereby enhancing the therapeutic efficacy and reducing side effects, playing a crucial role in ensuring the safety and accuracy of the treatment. With the development of accelerator technology, accelerators-based BNCT has entered the stage of rapid development. However, the development of TPS in China lags, significantly impeding the progress of BNCT clinical applications. In this paper, a program was designed independently using the Python language, which implemented the primary functions of BNCT-TPS, including medical data processing, organizational classification, three-dimensional reconstruction, particle transport simulation, dose distribution output, and data fusion. Interpolation functions and adjustable voxel grids were used to establish a model for patients to meet the requirements of simulation. The dose distribution of the model was simulated using MCNP, and further weighting and adjustments were made in the post-processing software based on the main elements affecting the efficacy of BNCT. Subsequently, the dose results and distribution maps of interest to physicians were overlaid and output with medical images using Python. After the completion of the entire program construction, a reactor neutron source was used to experimentally verify the simulation results of a 3D printed human head model. The distribution of neutron flux inside the human head was measured by activating gold foil and the simulation results were compared against the experimental results. The consistency of the outcomes confirms the validity of the data processing and material conversion method used in this paper. After verifying the function of the program, the dose distribution based on medical datasets of real human heads and neutron sources dedicated to BNCT were simulated. The simulation results show that the energy spectrum of the neutron source has a crucial impact on the therapeutic effect. The fast neutrons in the neutron source interact with the H element, resulting in a high dose on the skin and damaging the normal tissues of the body. The higher content of epithermal neutrons in the neutron source is beneficial for forming a higher thermal neutron fluence at the tumor, which increases the probability of interaction between boron-carrying drugs and thermal neutrons, thereby improving the therapeutic effect. This work establishes a basis for further development of a full-fledged TPS software.