Abstract: In the assembly process of carbon fiber reinforced resin matrix composite wall panels and frames, assembly gaps are likely to occur due to component manufacturing deviations and assembly errors. Traditional gap measurement methods, such as using feeler gauges, suffer from low accuracy, low efficiency, and difficulty measuring gaps in open or hard-to-reach areas. This paper proposed a gap prediction method based on actual measurement data. The external shape data of the components in their free state were obtained through digital measurement. The point cloud data were preprocessed, and the model was reconstructed. Finite element analysis was used to assess the shape changes of the wall panel under assembly constraints. Virtual assembly was conducted based on positioning references, and the gap sizes and distribution locations were calculated. The assembly gap prediction method proposed in this paper has been experimentally validated, and it demonstrates high accuracy in predicting the gaps between the wall panel and the frame. This method enables accurate gap predictions before actual assembly, helping to avoid repeated trials and improve efficiency.