This paper presents results of QSAR (Quantitative Structure Activity Relationship) studies realized with the PRECLAV (Property Evaluation by Class Variables) software. The database contains 66 derivatives of aromatic benzene sulfonamides incorporating 1, 3, 5-triazine moieties, fluorophenyl sulfamates, S-substituted-2-mercaptobenzenesulfonamide and diazenylbenzenesulfonamides with clinically used CA inhibitors. For each molecule over 3600 descriptors were calculated using programs MOPAC, PRECLAV and DRAGON. A heuristic algorithm selects the best multiple linear regression (MLR) equation showed that the correlation between the observed values and the calculated values of activity is very good (N = 66, Se = 0.263, r(2) = 0.884, F = 92.98, r(2)(cv) = 0.859, Q = 0.794). The virtual molecular fragments that lead to a significant increase of the inhibitor activity of hCA IX are C(3)H(2)N(5)Cl, NH(2), C(6)H(4), C(3)H(5)N(6), COOH, and C(3)HN(6). The virtual fragment--HO, C(5)H(2)NO, C(3)HN(6), leads to a significant decrease of the inhibitor activity value. With a view to external validation, the calibration set includes 50 molecules (Se = 0.256, r(2) = 0.885, F = 69.501, r(2)(cv) = 0.852) and the validation set includes 16 molecules (Se = 0.111, r(2) = 0.87, F = 93.984). Identification of molecules in validation set with high estimated value of inhibitory activity of hCA IX is correct enough to have practical value, even if the calibration/validation set contains aromatic benzene sulfonamides incorporating 1,3,5-triazine moieties and fluor phenyl sulfamates derivatives with very different chemical structures.