Abstract: Currently, the absence of cross-spectrum conditions in the two-stage vibration testing of aircraft complicates the selection and setting of coherence coefficients related to the cross-spectrum. This paper presented a theoretical analysis and experimental investigation to address this issue. The study analyzed the correlation between excitation sources during flight and ground vibration tests to elucidate the principles for determining coherence coefficients in testing. Through statistical analysis, the patterns of coherence coefficient deviations across frequency distributions and the reasons behind abnormal deviations in tests were identified. The impact of different coherence coefficients on response magnitude, equilibrum speed, and driving magnitude at measurement points was verified using a two-stage vibration test on a slender body spacecraft. Based on the analysis and test results, a method for selecting coherence coefficients for various frequency bands with moderate amplification tolerance was proposed. The experiment demonstrates the feasibility of the proposed method, and the findings can be applied to establish cross-spectra in multi-stage vibration tests.