RNA molecules play vital biological roles, and understanding their structures gives us crucial insights into their biological functions. Model evaluation is a necessary step for better prediction and design of 3D RNA structures. Knowledge-based statistical potential has been proved to be a powerful approach for evaluating models of protein tertiary structures. In present, several knowledge-based potentials have also been proposed to assess models of RNA 3D structures. However, further amelioration is required to rank near-native structures and pick out the native structure from near-native structures, which is crucial in the prediction of RNA tertiary structures. In this work, we built a novel RNA knowledge-based potential:PTRNAmark, which not only combines mutual and self energies but also fully considers the specificity of every RNA. The benchmarks on different testing data sets all show that PTRNAmark are more efficient than existing evaluation methods in recognizing native state from a pool of near-native states of RNAs as well as in ranking near-native states of RNA models.