Since the white layer thickness influences the surface quality of the machined specimens using electrical discharge machining process, the prediction of such parameter is highly important in the present scenario. Adaptive network based fuzzy inference system based white layer thickness prediction on machining processed silicon steel has been attempted in the present study. Three machining process parameters such as open circuit voltage, peak current and duty factor have been utilized for the training purpose owing their importance on determining white layer thickness. The accuracy of the prediction has been analyzed by comparing the predicted values from the architecture testing with the real time measured values. From the experimental results, it has been found that the developed adaptive network based fuzzy inference system can predict the average white layer thickness in an efficient way with accuracy of 96.8%. It has also been observed that the electrical process parameters have highly contributed on determining average white layer thickness. © 2019, Springer Nature B.V.