Cytomegalovirus is a leading cause of congenital disease and a vaccine is a high priority. The viral gB glycoprotein is essential for infection on all cell types. The guinea pig is the only small animal model for congenital CMV (cCMV), but requires guinea pig cytomegalovirus (GPCMV). Various GPCMV gB vaccine strategies have been investigated but not with a full length protein. Previous GPCMV gB vaccines have failed to fully protect against cCMV, with approximately 50% efficacy. In an effort to define the basis of GPCMV gB based vaccine failure, we evaluated recombinant defective Ad vectors encoding GPCMV gB full length (gBwt), or truncated protein lacking transmembrane domain (gBTMD). Both candidate vaccines evoked high anti-gB titers and neutralized virus infection on fibroblast cells but had varying weaker results on non-fibroblasts (renal epithelial and placental trophoblasts). Non-fibroblast cells are dependent upon the viral pentamer complex (PC) for endocytic pathway cell entry. In contrast, fibroblasts cells that express the viral receptor platelet derived growth factor receptor alpha (PDGFRA) to enable entry by direct cell fusion independent of the PC. Anti-gBwt sera was approximately 2-fold (renal epithelial) to 3-fold (fibroblasts) more effective at neutralizing virus compared to anti-gBTMD sera. Both gB vaccines were weakest against virus neutralization on trophoblasts. Knockout of PDGFRA cell receptor on fibroblast cells (GPKO) rendered virus dependent upon the PC pathway for cell entry and anti-gB GPCMV NA50 was more similar to epithelial cells. In a gBwt vaccine protection study, vaccination of animals significantly reduced, but did not prevent dissemination of wild type GPCMV challenge virus to target organs. Depletion of complement in vivo had limited impact on vaccine efficacy. Overall, a full length gB antigen has the potential to improve neutralizing antibody titer but fails to fully prevent virus dissemination and likely congenital infection.,Copyright © 2020 Elsevier Ltd. All rights reserved.