Human cytomegalovirus (CMV) is a significant infectious agent causing disease in immunocompromised HIV-infected patients, transplant recipients, and neonates. The current antiviral therapeutic strategy against CMV is limited in its utility due to the inherent toxicity and lack of bioavailability of currently available anti-CMV agents, ganciclovir (GCV), cidofovir (CDV), and foscarnet (FOS). The development of the prodrug of GCV, valganciclovir (val-GCV), has vastly improved the bioavailability profile of GCV. However, val-GCV demonstrates limited effectiveness against tissue-invasive CMV diseases as side effects involved with traditional intravenously administered GCV such as haematologic and reproductive toxicities remain. In addition, the emergence of antiviral resistant CMV mutant strains due to prolonged treatment with currently available antivirals necessitates the development of novel anti-CMV agents with reduced toxicity and improved bioavailability. In this study, select groups of novel compounds were analysed for their potential for further development as anti-CMV agents. Three groups of compounds were identified based on two screening methods which included the computer simulated screening process of compounds known as in silico screening and the traditional method of random screening. The first group of compounds (CATi) were identified by in silico screening against the CMV DNA polymerase catalytic aspartate triad, resulting in the identification of 31 compounds with the potential for inhibitory activity against CMV. The second group of compounds (PRO-i) were identified through in silico screening against the CMV protease, identifying a total of 18 lead compounds exhibiting structural complementarity with CMV protease. The third and final group of compounds (TPEX) were identified through random screening and consisted of plant extracts purified from tropical plants. All three compounds were initially screened for cytotoxicity against human fibroblasts. Plaque reduction assays were performed using compounds with acceptable levels of cytotoxicity to determine the ability of the compounds to inhibit the replication of the laboratory antiviral sensitive CMV strain, Towne. Two of the PRO-i compounds demonstrated good antiviral activity against CMV. Eleven percent (2/18) of the PRO-i compounds inhibited CMV replication, with PRO-i-43 and PRO-i­-44 displaying mean 50% inhibitory concentrations (IC50) of 4.8 ± 1.2 µM and 8.04 µM, respectively. PRO-i-43 and PRO-i-44 are thus good candidates for further development as novel antiviral agents against CMV. The majority of CATi and TPEX compounds displayed significant cytotoxicity against human fibroblasts and compounds with acceptable levels of cytotoxicities did not significantly inhibit CMV replication. However, the identification of compounds with low cytotoxicities provides a good foundation for further development of novel anti-CMV agents with superior antiviral activity. In silico screening against three-dimensional viral protein models is a useful strategy for the identification of novel antiviral agents with the potential for inhibitory activity against CMV. Structural modification to produce potent derivatives of the identified anti-CMV compounds (PRO-i-43 and PRO-i-44) is a good option for the further development of novel antiviral agents against CMV. Such further examination of the identified compounds with anti-CMV activity is required to investigate their activity against not only antiviral sensitive CMV strains but also resistant CMV strains. Further investigations will yield new insights into their target, allowing further identification of compounds with potential anti-CMV activity with pharmaceutical application.