Catalysis is of pivotal importance to many aspects of modern society, from chemicals synthesis to energy production and to pollutants remediation. Along with the growing world population and industrialization scale, rapid increases in global energy demands and environmental issues create a formidable challenge in designing new catalysts, which should be more active, more selective, more stable, and preferably comprised of earth-abundant elements. Traditional trial-and-error method no longer meets this fast-increasing requirement, which typically takes a long period for the basic research in materials design translating to manufacturing. Computational modelling can accelerate this process and greatly shorten the timescale. In this presentation, I will introduce our recent progress in catalysts development for oxygen reduction reaction (ORR) in fuel cell and Li-air battery, and oxidation reaction in diesel exhaust by integrating theory and experiment. Density functional theory (DFT) method is used to describe surface chemical reactions in atomic scale and to uncover the underlying principles that govern the catalytic activity. We further validate the theoretical predictions through experimental evidence and develop general descriptors for new catalysts design and optimization.