In this work, a thermal model based on finite element method has been proposed to predict the material removal and tool wear in electrical discharge machining process by simulating a single spark discharge. The model is validated by conducting experiments on a die-sinking EDM machine. It is found that numerical model provides reasonably accurate estimation of responses. Parametric analysis is conducted on the proposed model to study the influence of significant process parameters on the responses. Regression analysis is carried out to develop empirical models relating responses with process parameters. Finally, a non-dominated sorting genetic algorithm has been employed to obtain Pareto optimal solution for multi-objective optimisation. The numerical method provides an inexpensive and time saving alternative to study the performance of machining before actual cutting operation. Copyright © 2016 Inderscience Enterprises Ltd.