A central composite experimental design was used to investigate experimental process conditions (extraction temperature, extraction time and biomass loading) for the supercritical ethanol extraction of bio-oils from German beech wood. The effects of three independent variables and their interactions on the yields of ethanol extracts and biomass conversion were analyzed using response surface methodology (RSM). The most significant variable was found to be the temperature on the yields of ethanol extract and biomass conversion. High correlation coefficients (R-2) were obtained and they were 0.93 and 0.95 for ethanol extract and biomass conversion yields, respectively. The consistent matching of predicted values with the experimental values is an indication of validity of the model used in this study. Optimized conditions were in the range of the experimental design which shows the validity of the design. Optimal conditions for the ethanol extract yield were 315.81 degrees C extraction temperature, 123.67 min extraction time, and 15.95 wt% biomass loading. The bio-oil contained mainly phenolic compounds. The analyses of results show that the linear coefficient of x(1) and the quadratic coefficient of x(1)(2) are the most statistically significant factors on the ethanol extract and biomass conversion yield. The ethanol extracts (bio-oils) contained greater amounts of carbon and lesser amounts of oxygen than that of the raw material. (C) 2013 Elsevier B.V. All rights reserved.