The purpose of this article is to consider some recent developments in the area of the computational chemistry of metal–organic frameworks (MOFs), and more specifically on their crystal structure prediction and electronic structures. We intend here to illustrate how computational approaches might be powerful tool for the discovery of new families of hybrid frameworks, helping to understand their often complex energy landscapes. Also, MOFs have attracted a lot of attention due to their potential use for photocatalysis and optoelectronic, making it necessary to develop strategies to control their electronic structures. We will show how recent computational studies in this area have allowed a better understanding of their electronic properties and their potential tunability, highlighting when they have given successful guidelines for the discovery of novel MOFs with targeted properties.