Conceptually, blocking human immunodeficiency virus type 1 (HIV-1) integration is the last possibility for preventing irreversible cellular infection. Using cocultures of monocyte-derived dendritic cells and CD4+ T cells, which represent primary targets in sexual transmission, we demonstrated that blocking integration with integrase strand transfer inhibitors (InSTIs), particularly L-870812, could consistently block cell-free and cell-associated HIV-1 infection. In a pretreatment setting in which the compound was present before and during infection and was afterwards gradually diluted during the culture period, the naphthyridine carboxamide L-870812 blocked infection with the cell-free and cell-associated HIV-1 Ba-L strain at concentrations of, respectively, 1,000 and 10,000 nM. The potency of L-870812 was similar to that of the nucleotide reverse transcriptase inhibitor R-9-(2-phosphonylmethoxypropyl) adenine (PMPA) but one or two orders of magnitude lower than those of the nonnucleoside reverse transcriptase inhibitors UC781 and TMC120. In contrast, the diketo acid RDS derivative InSTIs showed clear-cut but weaker antiviral activity than L-870812. Moreover, L-870812 completely blocked subtype C and CRFO2_AG primary isolates, which are prevalent in the African heterosexual epidemic. Furthermore, the addition of micromolar concentrations of L-870812 even 24 h after infection could still block both cell-free and cell-associated Ba-L, opening the prospect of postexposure prophylaxis. Finally, an evaluation of the combined activity of L-870812 with either T20, zidovudine, PMPA, UC781, or TMC120 against replication-deficient HIV-1 Ba-L (env) pseudovirus suggested synergistic activity for all combinations. Importantly, compounds selected for the study by using the coculture model were devoid of acute or delayed cytotoxic effects at HIV-blocking concentrations. Therefore, these findings provide evidence supporting consideration of HIV-1 integration as a target for microbicide development.