To aid in the development of advanced ceramics, a materials design system was constructed. The system can be used to improve properties of ceramics by changing the composition and processing parameters. The system includes the following modules: the advanced ceramics database, the evaluation and optimization modules, the knowledge base, and the artificial neural network. Starting with the desired properties, a ceramics database will be sought. If suitable materials can be found, an evaluation module can be used to select the best one. If a suitable material is not found, the optimization module can be used to select a material with the desired property values as a reference material for materials design. The materials design system needs a large knowledge base. This knowledge base includes many relationships between composition, processing parameters, and properties. The traditional expert system or artificial neural network can be employed to acquire the knowledge base. Compared to traditional methods, the neural network is a more efficient tool for multidimensional, complex, and quantitative problems, especially problems with unknown rules. A combination of these two methods works best. Using the knowledge base obtained, users can predict the experimental results.