Abstract: The normal coefficient of restitution (R_n) is the most critical input parameter in the prediction and analysis of rockfall disasters. Because of its many controlling factors, how to determine an accurate and reasonable value of R_n is still a difficult problem. In order to explore the law and mechanism of the joint influence of multiple factors on R_n, the response surface methodology-central composite design method is used to investigate the combined effect of seven factors on R_n for the first time. Through regression and variance analysis of the test results, the seven selected factors have the significant effects on R_n, and the order of significance is the impact angle (θ) > block hardness (H₁) > rotational speed (ω) > shape factor (η) > incident velocity (V) > slope hardness (H₂) > size (d). Many interaction parameters also show the significant effects on R_n, in the order of d−θ > H₁−d > V−η > H₁−ω > d−η > η−ω > H₁−η. For a single factor parameter, R_n increases with the decrease of d, V, η, θ and the increase of H₁, H₂, ω; for interaction parameters, d and ω have a weak effect on R_n, only when H₁ is small, and the effect gradually increases with the increase of H₁; η has a significant effect on R_n when H₁ is small, and the effect significantly decreases with the increase of H₁; when η is small, R_n increases significantly with the decrease of V or d, and this effect is not significant when η is large; when θ is small, R_n decreases rapidly with the increase of θ, and the decreasing trend slows down when θ is large. Compared with block with small η, the ω of block with larger η has a more significant effect on R_n. These conclusions can provide an important reference for the construction of the R_n calculation model of rockfall, and provide a basis for the accurate prediction and prevention of rockfall disasters.