ABSTRACT: The effects of yeast extract (X1), NH4NO3 (X2), peptone (X3), urea (X4), CMC (X5), Tween 20 (X6), MgSO4 (X7), and CaCO3 (X8) on production of xylanase from Cellulosimicrobium cellulans CKMX1 were optimized by statistical analysis using response surface methodology (RSM). The RSM was used to optimize xylanase production by implementing the Central composite design. Statistical analysis of the results showed that the linear, interaction and quadric terms of these variables had significant effects. However, only the linear effect of X4, X5, interaction effect of X1X7, X1X8, X2X3, X2X8, X3X6, X3X8, X4X6, X4X7, X5X7, X5X8 and quadratic effect of X 32 , X 52 and X 72 found to be insignificant terms in the quadratic model and had no response at significant level. The minimum and maximum xylanase production obtained was 331.50 U/g DBP and 1027.65 U/g DBP, respectively. The highest xylanase activity was obtained from Run No. 30, which consisted of yeast extract (X1), 1.00 g (%); NH4NO3 (X2), 0.20 g (%); peptone (X3), 1.00 g (%); urea (X4), 10 mg (%); CMC (X5), 1.00 g (%); Tween 20 (X6), 0.02 mL (%); CaCO3 (X7), 0.50 g (%) and MgSO4 (X8), 9.0 g (%). The optimization resulted in 3.1-fold increase of xylanase production, compared with the lowest xylanase production of 331.50 U/g DBP after 72 h of incubation in stationary flask experiment. Application of cellulase-free xylanase in pulp biobleaching from C. cellulans CKMX1 under C-EP-D sequence has been shown to bring about a 12.5 % reduction of chlorine, decrease of 0.8 kappa points (40 %), and gain in brightness was 1.42 % ISO points in 0.5 % enzyme treated pulp as compared to control.