Ammonia is either a crucial resource of fertilizer production for solving the food problem of mankind or an important energy source as both an eco-friendly hydrogen carrier and a carbon-free fuel. Therefore, nowadays ammonia synthesis and decomposition become promising. Then, a catalyst is required to effectively perform the ammonia synthesis and decomposition. In order to design high-performing as well as cheap novel catalysts for ammonia synthesis and decomposition, it is necessary to test huge amount of catalyst candidates, but it is inevitably time-consuming and expensive to search and analyze using only traditional approaches. Recently, new methods using machine learning which is one of the core technologies of the 4th industrial revolution that can quickly and accurately search high-performance catalysts has been emerging. In this paper, we investigate reaction mechanisms of ammonia synthesis and decomposition, and we described recent research and prospects of machine learning-driven methods that can efficiently find high-performing and economical catalysts for ammonia synthesis and decomposition.