ABSTRACT Morphogenesis of herpesvirus particles is highly conserved; however, the capsid assembly and genome packaging of human cytomegalovirus (HCMV) exhibit unique features. Examples of these include the essential role of the small capsid protein (SCP) and the existence of the β-herpesvirus-specific capsid-associated protein pp150. SCP and pp150, as well as the UL77 and UL93 proteins, are important capsid constituents, yet their precise mechanism of action is elusive. Here, we analyzed how deletion of the open reading frames (ORFs) encoding pUL77, pUL93, pp150, or SCP affects the protein composition of nuclear capsids. This was achieved by generating HCMV genomes lacking the respective genes, combined with a highly efficient transfection technique that allowed us to directly analyze these mutants in transfected cells. While no obvious effects were observed when pUL77, pUL93, or pp150 was missing, the absence of SCP impeded capsid assembly due to strongly reduced amounts of major capsid protein (MCP). Vice versa, when MCP was lacking, SCP became undetectable, indicating a mutual dependence of SCP and MCP for establishing appropriate protein levels. The SCP domain mediating stable MCP levels could be narrowed down to a C-terminal helix known to convey MCP binding. Interestingly, an SCP-EGFP (enhanced green fluorescent protein) fusion protein which also impaired the production of infectious progeny acted in a different manner, as capsid assembly was not abolished; however, SCP-EGFP-harboring capsids were devoid of DNA and trapped in paracrystalline nuclear structures. These results indicate that SCP is essential in HCMV because of its impact on MCP levels and reveal SCP as a potential target for antiviral inhibitors. Human cytomegalovirus (HCMV) is a ubiquitous pathogen causing life-threatening disease in immunocompromised individuals. Virus-specific processes such as capsid assembly and genome packaging can be exploited to design new antiviral strategies. Here, we report on a novel function of the HCMV small capsid protein (SCP), namely, ensuring stable levels of major capsid protein (MCP), thereby governing capsid assembly. Furthermore, we discovered a mutual dependence of the small and major capsid proteins to guarantee appropriate levels of the other respective protein and were able to pin down the SCP domain responsible for this effect to a region previously shown to mediate binding to the major capsid protein. In summary, our data contribute to the understanding of how SCP plays an essential part in the HCMV infection cycle. Moreover, disrupting the SCP-MCP interface may provide a starting point for the development of novel antiviral drugs.