Monocyte/macrophages (M/M) are important targets for HIV in the body, and represent the majority of cells infected by the virus in some body compartments such as the central nervous system (CNS). M/M can be different from T-lymphocytes in terms of surface antigens, cell replication and drug metabolism. Thus, we evaluated, in M/M and in T-lymphocytes, the pattern of viral inhibition induced by various anti-HIV drugs, and assessed some of the mechanisms of action related to such antiviral activity. Inhibitors of HIV binding on CD4 receptors have similar activity in M/M and T-lymphocytes, while AZT and other dideoxynucleosides (ddN) are in general more active against HIV in M/M than in T-lymphocytes. This phenomenon can be related to the increased ratio in M/M of ddN-triphosphate/deoxynucleoside-triphosphate, and can at least in part explain the ability of zidovudine and didanosine in improving neurological dysfunctions in AIDS patients. Moreover, the antiviral activity of AZT (but not of other ddN- or HIV-binding inhibitors) is potently enhanced by cytokines like granulocyte-macrophage colony stimulating factor (GM-CSF) in M/M, while anti-HIV activity of TIBO compounds in M/M is not down-modulated by GM-CSF and other cytokines. Finally, non-toxic concentrations of adriamycin, an anticancer drug reported to be active against DNA viruses, can inhibit HIV replication in M/M (but not in T-lymphocytes). Taken together, these results suggest that M/M are selective targets for HIV with peculiarities different from those of T-lymphocytes. Thus, promising anti-HIV compounds should be evaluated both in T-cells and in M/M before reaching clinical trials. This may help in selecting drugs with good chances of being effective in patients with HIV-related disease.