Response surface methodology (RSM) and a five-level three-factor central composite rotatable design (CCRD) were used to evaluate the effect of glycerol and peptone concentration and initial pH on dextran dextrinase (DDase) production by Gluconobacter oxydans. Optimal fermentation conditions were 20.59 g/l of glycerol, 6.67 g/l of mycological peptone and an initial pH of 6.14. The predicted DDase yield of the optimised fermentation was 0.207 U/ml, whereas an actual experimental yield of 0.208 +/- 0.025 U/ml was obtained.