<p>During the late 20th century, evolutionary biology contributed to the notion that much of the genome is nonfunctional. This particularly applied to endogenous retroviruses (ERVs), which were seen as parasitic elements. However, growing evidence shows that ERVs play diverse and important roles, especially in gene regulation, development, and immunity. This article examines whether such findings pose a challenge to evolutionary biology, using Imre Lakatos’ model of scientific growth, which judges research programs by their ability to direct fruitful research and anticipate new data. I argue that the current number of known functional ERVs does not significantly contradict the evolutionary program’s low expectations, but that the degree of functionality will rise to anomalous levels if the recent trend of new discoveries continues. Moreover, the evolutionary program has contributed little to the discovery of the functions that have been found so far, except for a few cases where functionality was implied by sequence conservation. Nevertheless, evolutionary biology does possess the theoretical tools to retroactively produce mechanistic explanations for the origin of functional ERVs. The most common type of function—tissue-specific regulation of nearby genes—can be neatly explained in a neo-Darwinian fashion if the newly inserted ERVs immediately exerted a beneficial effect on host gene expression. This model predicts, often successfully, that the relevant transcription factor binding sites are present in the family’s consensus sequence, but has been criticized because ERV insertions with strong regulatory effects are nearly always deleterious. Another notable ERV function is facilitating cell–cell fusion in the placenta. Here, the program is forced to invoke the independent origin of ERV-derived fusion proteins in different mammalian lineages. Taken together, this evaluation yields an ambivalent verdict on the success of the evolutionary research program.</p>