Reactive intermediates play a critical role in the chemist's understanding of many reaction mechanisms. 1 Given their often fleeting appearance, experimental observation, let alone acquiring any property data on them, can be a great challenge. It is not, therefore, surprising that computational chemistry has been a strong partner with experiment in gaining knowledge of a broad spectrum of reactive intermediates.In this chapter, we focus on the class of reactive intermediates that bear at least two unpaired electrons: diradicals and carbenes. The exact definition of a diradical is somewhat in the eye of the beholder. Salem and Rowland 2 provided perhaps the most general, yet effective, definition-a diradical is a molecule that has two degenerate or nearly degenerate orbitals occupied by two electrons. With this definition, carbenes can be considered as a subcategory of diradicals. In a carbene, the two degenerate molecular orbitals are localized about a single carbon atom.The chemistry of radicals has been described many times. 3 -6 Likewise, the chemistry of diradicals and carbenes has been the subject of many reviews. 7 -16 Two issues will guide our presentation in this chapter. First and foremost, we discuss examples of diradicals and carbenes, where computational chemistry has greatly aided in understanding the properties, structure, and chemistry of these intermediates. A theme that will emerge here is the strong collaborative relationship between experimental and computational chemists that greatly aided in resolving the controversies and discrepancies that arose in trying to understand these unusual species. The second aspect of this chapter is that computational studies of diradicals and (especially) carbenes helped to establish the discipline of computational chemistry. In fact, the first example we discuss in this chapter is the nature of methylene, which is the topic that many claim to have single handedly won over skeptics to the advantages and power of computational chemistry.Following the section on methylene, we present the chemistry of phenylcarbene and phenylnitrene and describe how computational chemistry helped detail why these two closely related molecules behave so differently. A discussion of tetramethyleneethane (TME) and oxyallyl diradicals explores how theories of apparently simple molecules may be quite complicated. Next, we discuss the chemistry of Computational Organic Chemistry, Second Edition. Steven M. Bachrach