Abstract: A brief review of the early work, current research areas, and future development of computational electromagnetics in the computational electromagnetics laboratory (CEMLAB) at University of Electronic Science and Technology of China (UESTC) is described. The time-domain and frequency-domain numerical methods are developed to improve simulation efficiency while maintaining the same accuracy for transmission, radiation and scattering problems. The neural network with knowledge-based neurons and generalized transmission line equation techniques are adopted for electromagnetic modeling as CAD tools with high performance. Efficient electromagnetic optimization methods, such as jumping genes genetic algorithm and space mapping technique, are also proposed to optimize antennas and microwave passive components.