In addition to the interplay of the retrovirus and a particular cell, it is important to bear in mind that retroviruses have evolved in the context of the whole organism, and their behavior should be considered in this context. Thus, events, such as integration at a specific site, that are rare at a single-cell level are essentially certain to occur in some cells in an infected animal. In a multicellular host, the consequences of viral infection arise via the effects of viral replication— which causes the destruction or transformation of specific populations of cells in the host. In addition, for retroviral infection to persist, the virus must be able to evade or ignore the host's immune system. Finally, viruses must provide for their transmission from one host animal to another. For the virus to survive, an infected host must transmit it, on average, to at least one new host.

Unlike most RNA viruses, retroviruses establish lifelong infections. In most cases, the “natural” host-virus interaction appears to be benign, with the virus having little, if any, pathogenic effect during the natural lifespan of the host. Since animal-to-animal transmission of retroviruses is usually quite inefficient, the evolutionary survival of these viruses has depended both on establishing a lifelong infection and on allowing the infected animal to live long enough to transmit the virus. Pathogenic effects of retroviral infections seem to be a consequence of viral replication in an unnatural host (like HIV in humans) where the normal balance between replication and damage to the host is lost, of allowing animals to live for prolonged periods in a laboratory setting (where their lifespan is longer than it would be in the wild), or of selection in the laboratory of special variants, such as oncogene-containing viruses, which occur only very rarely in nature. Unlike viruses such as influenza or rabies, there is no evidence that retroviral pathogenesis directly aids transmission.