HIV-1 capsid (CA) plays indispensable and multiple roles in the life cycle of HIV-1, become an attractive target in antiviral therapy. Herein, we report the design, synthesis, and mechanism study of a novel series of dimerized phenylalanine derivatives as HIV-1 capsid inhibitors using 2-piperazineone or 2,5-piperazinedione as a linker. The structure-activity relationship (SAR) indicated that dimerized phenylalanines were more potent than monomers of the same chemotype. Further, the inclusion of fluorine substituted phenylalanine and methoxyl substituted aniline was found to be beneficial for antiviral activity. From the synthesized series, Q-c4 was found to be the most potent compound with an EC<sub>50</sub> value of 0.57 μM, comparable to PF74. Interestingly, Q-c4 demonstrated a slightly higher affinity to the CA monomer than the CA hexamer, commensurate with its more significant effect in the late-stage of the HIV-1 lifecycle. Competitive SPR experiments with peptides from CPSF6 and NUP153 revealed that Q-c4 binds to the interprotomer pocket of hexameric CA as designed. Single-round infection assays showed that Q-c4 interferes with the HIV-1 life cycle in a dual-stage manner, affecting both pre-and post-integration. Stability assays in human plasma and human liver microsomes indicated that although Q-c4 has improved stability over PF74, this kind of inhibitor still requires further optimization. And the results of the online molinspiration software predicted that Q-c4 has desirable physicochemical properties but some properties still have some violation from the Lipinski rule of five. Overall, the dimerized phenylalanines are promising novel platforms for developing future HIV-1 CA inhibitors with considerable potential for optimization.