Response surface methodology was employed to optimize medium composition for the production of cyclic adenosine 3',5'-monophosphate (cAMP) with Microbacterium sp. no. 205. A fractional factorial design (2(11-7)) was applied to evaluate the effects of different components in the medium. K(2)HPO(4), MgSO(4) and NaF were found to significantly influence on the cAMP production. The steepest ascent method was used to access the optimal region of the medium composition. The concentrations of the three factors were optimized subsequently using central composite design and response surface methodology. The optimal medium composition to achieve the optimal cAMP production was determined (g/L): K(2)HPO(4), 12.78; MgSO(4), 3.53 and NaF, 0.18. The corresponding cAMP concentration was 8.50 g/L, which was about 1.8-fold increase compared with that using the original medium. Validation experiments were also carried out to prove the adequacy and the accuracy of the model obtained. The cAMP fermentation in 5L fermenter reached 9.87 g/L.