To address the shortcomings that the existing assignment methods in the target threat assessment process can hardly reflect the contribution of indicators and the threat level classification is too subjective, a target threat level assessment method based on random forest assignment and TOPSIS\|RSR is proposed. The method first establishes a random forest model and determines the indicator weights by calculating the correlation between the indicators and the threat values in the model. Then a threat rank assessment model based on TOPSIS\|RSR was constructed to obtain the threat ranking and threat level of the target by ranking the relative superiority and inferiority distances, and the reasonableness of the threat ranking results was tested by ANOVA analysis. Finally, the effectiveness of the proposed method is proved by simulation experiments and comparative analysis.