Two novel classes of pyrrolobenzothiazepinones and pyrrolobenzoxazepinones were investigated as potential anti-AIDS drugs. These compounds were found to inhibit HIV-1 reverse transcriptase (RT) enzyme in vitro and to prevent HIV-1 cytopathogenicity in T4 lymphocytes, without appreciable activity on HIV-2 cytopathic effects, and against HBV as well as calfthymus DNA alpha-polymerase. Their potency is influenced by substituents at position 6 and on the fused aromatic ring. Specifically, small lipophilic substituents at C-6 were preferred, whereas substitutions on the benzo-fused ring were found to be detrimental to activity, with respect to the unsubstituted compounds. Modification of the pie-system at C-6 is well tolerated, although the replacement of the benzo-fused with a [2,3]naphtho-fused ring leads to a less active compound. Maximum potency and specificity is achieved with a phenyl and an ethyl group at position 6 of the pyrrolobenzoxazepinone system. In the enzymatic assay the oxazepinone derivative (+/-)-6-ethyl-6-phenylpyrrolo[2,1-d][1,5] benzoxazepin-7(6H)-one 16e (IC50 = 0.25 microM) was found to be more potent than nevirapine (IC50 = 0.5 microM), tested in the same experimental conditions using rC.dG as a template-primer. In cell culture assay benzoxazepine 16e was active against HIV-1, both wild type and AZT-sensitive, and HIV-1 (IIIB) strains, but not against HIV-2. In enzyme assay although 16e inhibited HIV-1 RT, it was inactive against the nevirapine-resistant recombinant RT Y181C at 50 microM. Molecular modeling studies suggest that these derivatives present a 3D pharmacophoric arrangement similar to that of other non-nucleoside inhibitors such as nevirapine.