Spatial resolution is one of the critical metrics for evaluating the performance of a computed tomography (CT) system. Traditional methods often neglected the influence of the focal spot size of the X-ray source, leading to data inconsistency and degrading the spatial resolution of the reconstructed images. Thus, this study introduces what we believe to be a novel non-linear super-resolution CT reconstruction method based on the characteristics of the X-ray source’s focal spot. The proposed method employed a discrete focal spot model and utilized measured focal spot information to formulate a non-linear mathematical model for CT imaging. Building on this model, a high-precision iterative solution method was developed. The proposed approach achieved improved data consistency during the forward projection process and employed a highly accurate solution method in the inversion process. As a result, this approach reconstructed images of higher quality compared to other methods, revealing more detailed structural information.