In order to develop a procedure for production bioethanol from cassava pulp, mixed culture of Amylomyces rouxii TISTR 3667 with Zygosaccharomyces pseudorouxii TISTR 5966 or Zymomonas mobilis TISTR 550 and cellulase were evaluated. The parameters such as pH, cellulase, and cassava pulp concentration that influence on the amount of fermentable sugar were optimized by response surface methodology (RSM). Ethanol production was observed and compared to the predicted value that was calculated from the models. The models were fitted to a second-order polynomial equation. They were used to predict ethanol concentration from the use of the mixed culture of A. rouxii TISTR 3667 and Z. mobilis TISTR 550 (G2) that was higher than the amount produced using the mixed culture of A. rouxii TISTR 3667 and Zygosaccharomyces pseudorouxii TISTR 5966 (G1). The following optimum parameters were obtained: pH 6, 20% cassava pulp, and 1% cellulase for G2; and pH 4, 20% cassava pulp, and 0.55% cellulase for G1. The effect of cellulase on ethanol production, a comparative study was conducted in the fermenter by using mixed culture of A. rouxii TISTR 3667 and Z. mobilis TISTR 550. It was showed that more 15% ethanol was gained from 10% cassava pulp with 0.5% cellulase (25 g/l ethanol) compared to the system without cellulase (20 g/l). Mathematically model (equation 4) predicted the ethanol in this system near the actual value of 26.87 g/l. This study indicated that RSM is a powerful tool for optimization fermentation process by using mixed culture including cellulase. Besides these cellulase also reduced viscosity of the cassava medium and enhanced ethanol production. However this process should be more continue to study.