Implicit with the use of animal models to test human cytomegalovirus (HCMV) vaccines is the assumption that the viral challenge of vaccinated animals reflects the anticipated virus-host interactions following exposure of vaccinated humans to HCMV. Variables of animal vaccine studies include the route of exposure to and the titer of challenge virus, as well as the genomic coding content of the challenge virus. This study was initiated to provide a better context for conducting vaccine trials with nonhuman primates by determining whether the in vivo phenotype of culture-passaged strains of rhesus cytomegalovirus (RhCMV) is comparable to that of wild-type RhCMV (RhCMV-WT), particularly in relation to the shedding of virus into bodily fluids and the potential for horizontal transmission. Results of this study demonstrate that two strains containing a full-length UL/b' region of the RhCMV genome, which encodes proteins involved in epithelial tropism and immune evasion, were persistently shed in large amounts in bodily fluids and horizontally transmitted, whereas a strain lacking a complete UL/b' region was not shed or transmitted to cagemates. Shedding patterns exhibited by strains encoding a complete UL/b' region were consistent with patterns observed in naturally infected monkeys, the majority of whom persistently shed high levels of virus in saliva for extended periods of time after seroconversion. Frequent viral shedding contributed to a high rate of infection, with RhCMV-infected monkeys transmitting virus to one naïve animal every 7 weeks after introduction of RhCMV-WT into an uninfected cohort. These results demonstrate that the RhCMV model can be designed to rigorously reflect the challenges facing HCMV vaccine trials, particularly those related to horizontal transmission.