Molecules of the structure ppp(A2'p)2A containing a 2' leads to 5' phosphodiester bond, commonly abbreviated as 2-5A, are synthesized in interferon-treated virally-infected cells and have been implicated in several systems as contributing to interferon's antiviral activity. The 2-5A binds to and subsequently activates an endogenous endonuclease, ultimately resulting in degradation of RNA. We have been interested in the use of 2-5A analogues to achieve antiviral activity without the use of interferon. For this approach to be successful, analogues must be synthesized with an increased stability (native 2-5A is rapidly degraded by cellular phosphodiesterases) and with increased ability to enter intact cells. Removal of the highly-negative charged 5' terminal phosphates from ppp(A2'p)2A results in formation of the 'core' species, (A2'p)2A, which should be able to penetrate intact cells more readily. While Kimchi et al. have shown that 2-5A core has an antimitogenic effect in mouse spleen lymphocytes and 3T3 fibroblasts, Williams and Kerr have reported lack of antiviral activity against Semliki Forest virus or encephalomyocarditis virus by exogenously-administered 2-5A core. We have previously determined that (xyloA2'p)2xyloA (abbreviated as xylo 2-5A core), the xyloadenosine analogue of the 5'-terminally dephosphorylated 2-5A core, is over 100 times more stable than the parent 2-5A core species. We now report that this xylo 2-5A core inhibits replication of herpes simplex viruses 1 and 2 in vitro, with greater than 100 times the activity of the parent 2-5A core. The mechanism of antiviral action of the 2-5A core analogue appears to involve a pathway different from that activated by the parent 5' triphosphorylated 2-5A species.