Response surface methodology was used to investigate the interaction of pH, temperature, and enzyme loadings on corn stover hydrolysis rates following soaking in aqueous ammonia pretreatment. Economic tradeoffs were estimated for cellulase and hemicellulase loadings under different hydrolysis conditions. Enzyme loadings had a more significant effect on rates than did pH or temperature. The effect of hydrolysis pH was independent of temperature and enzyme loadings, and the optimal pH for glucose and xylose yields were 4.5 and 4.3, respectively. Conducting hydrolysis at 50 degrees C rather than 37 degrees C enables either a 10% glucose yield increase, or a comparable yield with 40% and 65% reduction in cellulase and hemicellulase loadings, respectively. Although yield models showed that hydrolysis rates increase with higher enzyme loadings, economic models showed that optimal cellulase and hemicellulase loadings were as much as 47% and 23% lower, respectively, than the maximum loadings tested. Optimal enzyme loadings change with fluctuations in enzyme costs and ethanol price, but cellulase loadings were more sensitive to these changes than hemicellulase loadings. Enzyme loadings were also more sensitive to enzyme price at lower processing temperatures. Enzyme loadings can be adjuSted to increase return based on enzyme costs, ethanol price, and process temperature. (C) 2017 Elsevier Ltd. All rights reserved.