Background and Objective: The 3-mercaptopyruvate sulfurtransferase (3-MST) is an enzyme known to catalyse the detoxification of cyanide by transferring sulfur from appropriate substrates to cyanide to form a less toxic compound. Aqueous two-phase partitioning system (ATPS) has several advantages as compared with conventional methods of enzyme purification. However, to improve enzyme purification through ATPS, a combination of process variables need to be optimized. Materials and Methods: In this study, response surface methodology (RSM) was adopted to determine the optimum condition for the aqueous two-phase partitioning of chicken liver 3-MST. The variables optimized are polyethylene glycol (PEG), ammonium sulphate and NaCl. Results: The results obtained validate the predictability of the model. The optimal concentration of independent variables that ensued highest purification yield of 80.989% was 25% (PEG), 6.9% (ammonium sulphate) and 2.8% of NaCl. The R 2 value for the model was 0.9708 (p<0.05). Conclusion: The experimental values obtained in this study are following those predicted, indicating the suitability of the employed model and the success of the response surface methodology (RSM) in optimizing the purification conditions.