Test on a small scaled model is an effective approach to predict the dynamic response of full scale structure under blast loadings. However, the geometric dimensions of specimens cannot simply comply with complete geometrical similarity due to manufacture or test restrictions. It would result in the difference structural performance between the full and small scaled models. This paper proposed a corrected similarity relationship of the dynamic behaviour between prototype and replica of stiffened plates subjected to blast load, in which both the thickness of the plate and the configuration (cross-sectional shape) of stiffeners are distortedly scaled-down (double distorted geometric scaling factors). Firstly, based on the mesh convergence study and comparing with results from experimental tests, a numerical method in predicting the confined blast load and dynamic response of structure was verified, which provides a reliable means to determine the dynamic behaviour of stiffened plate designed by the corrected similarity criterion of this paper. Then, the influence of altering the stiffener configuration on the dynamic response of stiffened plates was analysed and on the basis of it, a criterion for scaling the stiffener is proposed to help design a stiffener-distorted model from prototype structure. In addition, a method for scaling the double-parameter distortedly small scaled model is proposed to predict the dynamic response of the prototype. Finally, two sets of examples of both the small size and prototype stiffened structures subjected to blast load were analysed by using the presented method. It is shown that the replica developed by applying the present method is able to accurately predict the behaviour of the full-size stiffened plates, even when the thickness of the plate and the configuration of the stiffeners are distortedly scaling down with different factors.