Microorganisms are increasingly being used in cyanide bioremediation. Several organisms have been reported to thrive in cyanide contaminated wastewater due to their ability to produce cyanide detoxifying enzymes. However, to improve the production efficiency of these enzymes combinations of process variables need to be optimized. In this study, Klebsiella oxytoca JCM 1665 was isolated from industrial wastewater, identified by sequencing its 16S rRNA gene and subjected to rhodanese production using submerged fermentation. The conditions for production were optimized using response surface methodology (RSM). Central composite design was employed to evaluate the effects of three production parameters – peptone (1 – 5 %), KCN (0.1 – 0.5 %), and time of incubation (1 – 24 h). Second-order polynomial model was used to predict the response. Rhodanese activity in the experiments varied from 0.05 to 7.5 RU.mg-1. Under the optimum conditions of 4.35 % peptone, 0.4 % KCN and incubation time of 13 hr., the value for rhodanese yield was 7.810 U.mL-1. The R2 value for the model was 0.9925 (R2 = 0.9925). Also, the experimental values are in accordance with those predicted, indicating the suitability of the employed model and the success of RSM in optimizing the production conditions.