A population replacement experiment has been devised to test the ability of a challenge virus to replace the resident virus in a persistently infected cell culture. BHK-21 cells persistently infected with foot-and-mouth disease virus of serotype C (clone C-S8c1) were challenged with a large excess of either the parental foot-and-mouth disease virus C-S8c1, genetically marked variants differing in their degree of virulence, or a mutant rescued after prolonged persistence in BHK-21 cells. After challenge, the composition of the resident virus population in the carrier culture was analyzed by reverse transcription-PCR amplification and nucleotide sequencing. The dominance of the initial persisting virus was seen in all cases, except when virulent viruses were used in the challenge. The experiments document that, paradoxically, virulence can be a positive factor in the reestablishment of a virus population in a persistently infected cell culture. A model based on the selection of virus-resistant cell variants during persistence is proposed to interpret these observations. Implications about the persistence of viruses in their host cells and organisms are discussed.