Human herpesvirus 6A (HHV-6A) and 6B (HHV-6B) are ubiquitous betaherpesviruses that infects humans within the first years of life and establishes latency in various cell types. Both viruses can integrate their genomes into telomeres of host chromosomes in latently infected cells. The molecular mechanism of viral integration remains elusive. Intriguingly, HHV-6A, HHV-6B and several other herpesviruses harbor arrays of telomeric repeats (TMR) identical to human telomere sequences at the ends of their genomes. The HHV-6A and HHV-6B genomes harbor two TMR arrays, the perfect TMR (pTMR) and the imperfect TMR (impTMR). To determine if the TMR are involved in virus integration, we deleted both pTMR and impTMR in the HHV-6A genome. Upon reconstitution, the TMR mutant virus replicated comparable to wild type (wt) virus, indicating that the TMR are not essential for HHV-6A replication. To assess the integration properties of the recombinant viruses, we established an in vitro integration system that allows assessment of integration efficiency and genome maintenance in latently infected cells. Integration of HHV-6A was severely impaired in the absence of the TMR and the virus genome was lost rapidly, suggesting that integration is crucial for the maintenance of the virus genome. Individual deletion of the pTMR and impTMR revealed that the pTMR play the major role in HHV-6A integration, whereas the impTMR only make a minor contribution, allowing us to establish a model for HHV-6A integration. Taken together, our data shows that the HHV-6A TMR are dispensable for virus replication, but are crucial for integration and maintenance of the virus genome in latently infected cells.