The retroviruses are a very large group of viruses that infect invertebrates as well as vertebrates. Most of what we know about this group of viruses comes from studies of viruses that higher vertebrates. Hundreds have been studied and, although considerable divergences exist, they form a well-defined taxon. All are sufficiently similar to be classified as belonging to a single family, the Retroviridae. The family gets it name from the concept that these viruses use retrograde flow of information, from RNA to DNA, whereas the conventional flow of information in living organisms is from DNA to RNA.

The RTs of Retroviruses are the most highly conserved elements of these viruses and have been used to study the relationships among them. Lineage of fish viruses, now classified as members of the genus Epsilonretrovirus. Based on these sequence relationships, the Retroviruses that infect birds and mammals are classified into six genera. Of these, members of three genera are characterized as simple retroviruses, which encode only the genes gag, pro, pol, and env (and sometimes dut). The other three genera retroviruses of higher vertebrates, as well as the fish viruses, encode, in addition, regulatory genes that control their life cycle, and they are called complex Retroviruses. Notice that these regulatory genes independently entered the four different lineages of complex Retroviruses represented by the four different genera. Thus, recombination to acquire new functions has been an ongoing process in the Retroviruses. Also notice that the complex Retroviruses do not group together.

The Epsilonretroviruses are more closely related to the Gammaretroviruses, which are simple viruses, than they are to other complex Retroviruses, and the Deltaretroviruses, Lentiviruses, and Spumaviruses are not particularly closely related.

Members of the different genera differ in their structure as visualized in the electron microscope. The simple viruses were formerly classified on the basis of morphology into groups A, B, C, and D. The nucleocapsids of C-type viruses, now classified as Alpharetrovirus and Gammaretroviruses, assemble during budding, and the nucleocapsid is centrally located in the mature virion. The nucleocapsids of B-type and D-type viruses, now classified as Betaretroviruses, assemble before budding and the nucleocapsid is eccentrically located (type B) or bar shaped (type D) in the mature virion.

Eukaryotic genomes contain a very large number of genetic elements that are related to retroviral genomes. Retrotransposons encode RT and can move around within the genome by a process that uses reverse transcription and insertion, similar to what happens with Retroviruses. They are related to retroviruses but have no independent lives as viruses. Other elements in the eukaryotic genome contain additional Retrovirus-like genes and appear to have arisen by insertion of retroviral into the germ line at some time in the past. Some of these are still active, capable of giving rise to infectious Retroviruses, whereas others are defective. It is clear that this class of elements has been coevolving with the eukaryotes for a long period of time. The integrated copies of Retroviruses in the germline constitute a form of fossil record that allows us to trace the lineage of at least some retroviruses for 100 million years or longer. For other viruses, whether RNA or DNA, we can only trace ancestry for much shorter periods of time.