Chemical beneficiation route for improvement of coal quality was demonstrated in pilot scale to produce low ash (8-10%) clean coal with a yield more than 70% by demineralization of high ash content flotation tailings obtained in a typical Indian coal washery. A Box-Behnken based design of experiment (DOE) method was used to generate data for development of a model. This model has been used to portray the effect of variables on reduction of ash content of coal tailings subjected to chemical leaching process. In this study, the functional relationship between process parameters (reaction temperature and time as well as reagent concentration) and the ash content of the final product was mapped using a second order non-linear model equation. The regression analysis was carried out using STATISTICA software for evolving appropriate equation to develop a commensurate model. The model predicted values of product coal ash content were in good agreement with the experiment values (R-2 = 0.9671). Moreover, the main effects, two-way interaction effects as well as sensitivity of process parameters were investigated with the help of the model. Finally, model based optimization was carried out to identify the operating parameters (reaction temperature = 160 degrees C; reaction time = 60.6 min; reagent concentration = 17.6%) at which the product coal with minimum ash content could be produced. (C) 2015 Elsevier B.V. All rights reserved.