First generation biofuels are predominantly produced from corn, sugarcane and vegetable oil using conventional technology, which raises serious concerns on fuel versus food. The abundant and renewable lignocellulosic biomass provides a great potential for second generation biofuels production. However, cost-effectiveness of second generation biofuels production is significantly hindered by biomass accessibility, high cost of pretreatment and enzymes, and less efficient utilization of hemicellulose. Hemicellulose, which accounts for 20-30% of dried woody biomass, can be readily broken down to simple sugars (primarily xylose and mannose) and other fermentation inhibitors (furfural) during the fractionation process. Although significant amount of research work has been done on xylose fermentation development, the final results are not promising thus far. Therefore, development of value added chemicals from hemicellulose is highly promoted currently. Chemical conversion of hemicellulose to furfural and hydroxymethylfurfural (HMF) has attracted substantial amount of attention associated with biofuels production. However, the low yield and concentration of hemicellulose stream previously extracted from fractionation process is limited by traditional pulping or pretreatment process. Optimization of hemicellulose extraction and recovery during the fractionation process will be crucial to develop value added co-products from woody biomass. The objective for this study is to develop a furfural/HMF oriented fractionation process for value added co-products and biofuels production from underutilized forest biomass. The maximum recovery of hemicellulose sugars from underutilized forest biomass was optimized under the controlled parameters of temperature, acid concentration, liquor-to-wood ratio, and retention time.