The classical Chinese poetry is a remarkable form of art in traditional Chinese character. However, it is difficult for people who are unfamiliar with ancient Chinese to experience the artistic content of the poetry. In this study, a sonification scheme, Tx2Ms (Text-to-Music), is proposed to extract the poetry features between lines in verses; moreover, dynamics and interval relations are modeled to map those features to the movement of multi-dimensional musical elements such as durations. This conversion is based on poetry intonation and acoustic analysis of the pronunciations of poems; and a stochastic compositional algorithm is created by applying a Markov chain. In addition, the best pentatonic mode for a specific poem is recommended according to the formants analysis. Therefore, the sonification of classical Chinese poetry not only provides a novel way for people to appreciate Chinese poetry but also enriches the state of mind and imagery in the delivery process, and the experiment results show that the proposed system is successfully accepted by most people.