This paper describes TFS-based artificial neural network (TFS/ANN) approach for classification and prediction of pharmacological active classes of chemicals. Dopamine antagonists of 1.227 that interact with different type of receptors (Dl, D2, D3 and D4) were used for the training. The TFS/ANN successfully classified 89% of the drugs into their own active classes. Then, the trained model was used for predicting class unknown compounds. For the prediction set of 137 drugs that were not included in the training set, the TFS/ANN model predicted 111 (81%) drugs of them into Iheir own active classes correctly. TFS method was also employed for similar structure searching and identification of active molecular analogues. The TFS successfully identified structurally similar molecular analogues of our interest.