Culture conditions were optimized for the growth and carbonyl reductase production by a novel yeast strain Candida viswanathii. Response surface methodology was applied for the critical medium components (initial pH, mannitol, yeast extract and calcium chloride) identified earlier by one-factor-at-a-time approach. Central composite design was used for the optimization studies. Using this methodology, the optimal values for the concentration of mannitol, initial pH, yeast extract and calcium chloride were 1.9, 7.5, 1.6 and 4, respectively. This medium was projected to produce, theoretically, growth having an optical density of 1.1 (600 nm) and an enzyme activity of 81.5 U/ml. Using this optimized medium, an experimental growth of 1.1 OD (600 nm) and enzyme activity 80.9 U/ml verified the applied methodology. This approach for medium optimization led to an enhancement of the growth and enzyme activity by 1.3 and 2.3 times higher, respectively, as compared to the unoptimized media.