Stirred tank reactors are most commonly used both in the laboratory and industry. Particularly for bioreactors, the volumetric mass transfer coefficient (kLa) of oxygen is used as one of the important parameters for determining efficiencies of reactors and for successful scale‐up. A number of correlation methods have been previously developed to predict the kLa in stirred tank bioreactors. In the present work, we propose a new correlation for kLa based on a mathematical and statistical approach using Response Surface Methodology (RSM) based on Box‐Behnken design of experiments. This correlation includes the effect of various parameters such as impeller agitation rate (50–800 rpm), air flow rate (0.5–3.5 L/min), and temperature (10−40°C) for different impeller configurations (single and dual Rushton, pitched blade, and mixed turbines). It was observed that the kLa increases with increasing the parameters for all the impeller configurations studied. Among the operating parameters, the most significant variable impacting kLa was found to be agitation rate, followed by air flow rate, and temperature. The models developed using RSM successfully interpreted the experimental kLa and were further validated under other operating conditions. It was also found that, compared with conventional power‐law models, the RSM approach enables a more efficient correlation procedure and formulates simplified models with comparably high accuracy, suggesting that the RSM is promising for evaluation of oxygen mass transfer in stirred tank bioreactors. © 2018 American Institute of Chemical Engineers Environ Prog, 38: 387–401, 2019