Background: Congenital cytomegalovirus (CMV) infection is a major public health problem. Antiviral therapies administered during pregnancy might prevent vertical CMV transmission and disease in newborns, but these agents have not been evaluated in clinical trials. The guinea pig model of congenital CMV infection was therefore used to test the hypothesis that antiviral therapy, using the agent agent cyclic cidofovir (cHPMPC), could prevent congenital CMV infection. Results: Pregnant outbred Hartley guinea pigs were challenged in the early-third trimester with guinea pig CMV (GPCMV) and treated with placebo, or the antiviral agent, cyclic cidofovir. To optimize detection of vertical infection, an enhanced green fluorescent protein (eGFP)-tagged virus was employed. Compared to placebo, cyclic cidofovir-treated dams and pups had reduced mortality following GPCMV challenge. The magnitude of GPCMV-induced maternal and fetal mortality in this study was reduced from 5/25 animals in the placebo group to 0/21 animals in the treatment group (p = 0.05, Fisher's exact test). By viral culture assay, antiviral therapy was found to completely prevent GPCMV transmission to the fetus. In control pups, 5/19 (26%) were culture-positive for GPCMV, compared to 0/16 of pups in the cyclic cidofovir treatment group (p < 0.05, Fisher's exact test). Conclusion: Antiviral therapy with cyclic cidofovir improves pregnancy outcomes in guinea pigs, and eliminates congenital CMV infection, following viral challenge in the third trimester. This study also demonstrated that an eGFP-tagged recombinant virus, with the reporter gene inserted into a dispensable region of the viral genome, retained virulence, including the potential for congenital transmission, facilitating tissue culture-based detection of congenital infection. These observations provide support for clinical trials of antivirals for reduction of congenital CMV infection.