An analysis of all known human herpesviruses has not previously been reported on sperm from normal donors. Using an array-based detection method, we determined the cross-sectional frequency of human herpesviruses in semen from 198 Danish sperm donors. Fifty-five of the donors had at least one ejaculate that was positive for one or more human herpesvirus. Of these 27.3% (n = 15) had a double herpesvirus infection. If corrected for the presence of multiple ejaculates from some donors, the adjusted frequency of herpesviruses in semen was 27.2% with HSV-1 in 0.4%; HSV-2 in 0.1%; EBV in 6.3%; HCMV in 2.7%; HHV-6A/B in 13.5%; HHV-7 in 4.2%, whereas none of the samples had detectable VZV or HHV-8. Subsequently, we examined longitudinally data on ejaculates from 11 herpesvirus-positive donors. Serial analyses revealed that a donor who tested positive for herpesvirus at one time point did not necessarily remain positive over time. For the most frequently found herpesvirus, HHV-6A/B, we examined its association with sperm. For HHV-6A/B PCR-positive semen samples, HHV-6A/B could be detected on the sperm by flow cytometry. Conversely, PCR-negative semen samples were negative by flow cytometry. HHV-6B was shown to associate with sperm within minutes in a concentration dependent manner. Confocal microscopy demonstrated that HHV-6B associated with the sperm head, but only to sperm with an intact acrosome. Taken together, our data suggest that HHV-6A/B could be transported to the uterus via binding to the sperm acrosome. Moreover, we find a 10 times higher frequency of HHV-7 in semen from healthy individuals than previously detected. Further research is required to determine the potential risk of using herpesvirus-positive donor semen. Longitudinally analyses of ejaculate series indicate that implementation of quarantine for a donor shown to shed a herpesvirus is not a tenable solution.