Levulinic acid (LA) has been identified as a promising green, biomass derived platform chemical. Response surface analysis (RSA) with a four-factor-five-level central composite design (CCD) was applied to optimize the hydrolysis conditions for the conversion of bamboo (Phyllostachys Praecox f. preveynalis) shoot shell (BSS) to LA catalyzed with ionic liquid [C(4)mim]HSO4. The effects of four main reaction parameters including temperature, time, C-[C4mim]HSO4 (initial [C(4)mim]HSO4 concentration) and X-Bss (initial BSS intake) on the hydrolysis reaction for yield of LA were analyzed. A quadratic equation model for yield of LA was established and fitted to the data with an R-2 of 0.9868, and effects of main factors and their corresponding relationships were obtained with RSA. Model validation and results of CCD showed good correspondence between actual and predicted values. The analysis of variance (ANOVA) of the results indicated that the yield of LA in the range studied was significantly (P<0.05) affected by the four factors. The optimized reaction conditions were as follows: temperature of 145 degrees C, time of 103.8 min, C-[C4mim]HSO4 of 0.9 mol center dot L-1 and X-BSS of 2.04% (by mass), respectively. A high yield [(71 +/- 0.41)% (by mol), triplicate experiment] was obtained at the optimum conditions of temperature of 145 degrees C, time of 104 min, C-[C4mim]HSO4 of 0.9 mol center dot L-1 and X-Bss of 2% (by mass), which obtained from the real experiments, concurred with the model prediction [73.8% (by mol) based on available C-6 sugars in BSS or 17.9% (by mass) based on the mass of BSS], indicating that the model was adequate for the hydrolysis process.