Artificial neural network was used to predict the effects of operational parameters on coal desulfurization using peroxyacetic acid from microwave pretreated coal. Coal particle size (150–1125 μm), leaching temperature (25–85 °C), leaching time (0–120 min), microwave irradiation power (0–1000 W) and time (0–110 s) were used as inputs to the network. The outputs of the model were organic and inorganic sulfur reductions for 40 of the data sets. The GRNN artificial neural network with spread of 0.3 was used to estimate both organic and inorganic sulfur reduction from a combined database, which was established from microwave pretreatment and leaching experiments. Thirty-two data sets were used for training and eight data sets for testing. Simulated values obtained from the neural network, correspond closely to the experimental results. Satisfactory correlations of R2 = 0.99 and 0.97 were achieved during the testing stages of the prediction of inorganic and organic sulfur reductions respectively.