The development of safe and effective antiviral agents has been a slow process, largely because of the difficulty in distinguishing between virus and host functions; materials toxic to the virus are frequently harmful also to the host in which the agent resides. Recently, techniques which target nucleic acid sequences as a means of reducing gene expression have emerged. This antisense armamentarium includes ribozymes, RNA enzymes which cleave other RNA molecules in a sequence-specific manner. We wish to assess the ability of ribozymes to control animal virus infection. Reasoning that the viruses most vulnerable to ribozyme intervention will be those whose complete life cycle is based on RNA (with no DNA stage), we have begun to develop ribozymes directed toward lymphocytic choriomeningitis virus (LCMV), the prototype of the arenavirus family. Using ribozymes of the hammerhead variety, we have identified several sites on the LCMV genome which can be efficiently cleaved in trans. The efficiency of cleavage is site dependent, and we demonstrate that secondary structure at the target site can abolish ribozyme cleavage. Computer-assisted analysis indicates that much of the LCMV genome may be involved in base pairing, which may render it similarly resistant to ribozyme attack. The few remaining open regions of LCMV lack a GUC target site, on which most studies to date have relied. Here we show that AUC, CUC, and AUU are alternative sites which can be cleaved by trans-acting ribozymes. This finding is important given the aforementioned restriction of available sites, imposed by secondary structure.