Rice husk as a widely available lignocellulosic material was subjected to an alkaline pretreatment process. The alkaline pretreatment was carried out under various conditions. The influence of process parameters, such as pretreatment time, solid loading, and NaOH concentration, on the glucose and xylose yields were investigated by means of appropriate models. The maximum glucose and xylose yields obtained under optimum pretreatment condition were 68.82% and 53.77%, respectively. Response surface methodology (RSM) and artificial neural network were used to model the pretreatment processes. Both modeling methodologies were statistically compared by means of the coefficient of determination and relative mean square error. It was concluded that the artificial neural network shows a somewhat better performance compared to RSM.