The three-dimensional quantitative structure–activity relationships (3D-QSAR) of a series of HOE/BAY-793 analogs (C2-symmetric diol peptidomimetics), developed by Budt and co-workers [Bioorg. Med. Chem. 3 (1995) 559] as inhibitors of HIV-1 protease (HIV-PR), were studied using Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA). Theoretical active conformers for these peptidomimetics were generated, derived from modeled protease inhibitor complexes, in order to orient the compounds superposition and to afford a consistent alignment. The best CoMFA model (N = 27, q2 = 0.637, R2 = 0.991) showed contributions of the steric (45.7%) and electrostatic (54.3%) fields to the activity, while the best CoMSIA model (N = 27, q2 = 0.511, R2 = 0.987) showed contributions of the electrostatic (68.5%) and hydrogen bond donor (37.5%) fields. The models were also external validated using four compounds (test set) not included in the model generation process. The statistical parameters from both models indicate that the data are well fitted and have high predictive ability. Moreover, the resulting 3D CoMFA/CoMSIA contour maps provide useful guidance for designing highly active ligands. The CoMFA/CoMSIA models were also compared with previous 4D-QSAR models [E.F.F. da Cunha, M.G. Albuquerque, O.A.C. Antunes, R.B. de Alencastro, QSAR Comb. Sci. 24 (2005), 240–253.].