Predictive hologram quantitative structure activity relationship (HQSAR) models were developed for a series of arylsulfonamide compounds acting as specific 5-HT6 antagonists. A training set containing 48 compounds served to establish the model. The best HQSAR model was generated using atoms, bond, and connectivity as fragment distinction and 4-7 as fragment size showing cross-validated r2(q2) value of 0.702 and conventional r2 value of 0.971. The predictive ability of the model was validated by an external test set of 20 compounds giving satisfactory predictive r2 value of 0.678. The efficiency of HQSAR approach was further evidenced by the generation of predictive models for a training set containing 30 highly diverse, both specific and nonspecific 5-HT6 antagonists. The best HQSAR model for this training set was generated using atoms, bond, and connectivity as fragment distinction and 4-7 as fragment size showing cross-validated r2(q2) value of 0.693 and conventional r2 value of 0.923. This model was also validated by using an external test set of 10 compounds giving satisfactory predictive r2 value of 0.692. The contribution maps obtained from these models were used to explain the individual atomic contributions to the overall activity.