Arenaviruses are a major cause of hemorrhagic fevers endemic to Sub-Saharan Africa and South America, and thus a major public health and medical concern. The prototypic arenavirus lymphocytic choriomeningitis virus (LCMV) is widely used as a model system for studying persistent and acute infections, as well as for gaining an understanding of mammalian immune function. When originally characterized three decades ago, the LCMV isolate, Armstrong, which causes an acute infection in adult mice, was found to differ from the LCMV Clone 13 strain that causes a persistent infection by two amino acid changes, one within the virus surface glycoprotein (GP1: F260L) and the other within the virus L polymerase (K1076Q). Mutation F260L was considered solely responsible for the exceptionally strong binding affinity of Clone 13 (L at GP1 260) to its cellular receptor, α-dystroglycan, which among cells of the immune system is preferentially expressed on dendritic cells, and consequently, alters dendritic cell function leading to viral persistence. Recently, we noted a previously overlooked nucleotide difference between these two strains that results in an additional amino acid change in GP1, N176D. To investigate the potential contribution of this newly identified mutation to the Clone 13 phenotype, we used reverse-genetics approaches to generate recombinant LCM viruses with each of these individual mutations. Phenotypic characterization of these rLCMV showed that mutation F260L, but not N176D, in the GP1 of LCMV is essential for mediating the long-term persistence of Clone 13 infections. This work emphasizes the importance of subtle differences in viral strains that determine disease outcomes.