Surface roughness and metal removal rate are the prominent output responses that influence the quality and cost of production. Moreover, in this present investigation neural network modeling and Adaptive-neuro fuzzy inference system (ANFIS) are employed to predict the output responses with respect to a variety of cutting parameters in cylindrical grinding process. A full factorial experimental design is conducted on AISI 1040 steel; considering work speed, depth of cut and feed rate as the cutting parameters. In addition 8 sets of experiments are performed to validate the computational methods. Our comparative study on identifying a better prediction methodology illustrated ANFIS as a better technique with 91% accuracy, while depth of cut as the most influential parameter effecting output responses.