Direct synthesis of dimethyl ether has been statistically analyzed in terms of product distribution and the effect of operating conditions. The investigated catalyst system consists of a combination of methanol synthesis (CuO–ZnO–Al2O3) and methanol dehydration (γ-Al2O3). The range of operating conditions varied as: T = 200–260 0C, H2/CO = 0.67-2 and SV (Space Velocity) = 41.74-292.68 h− 1 at P = 5.1 MPa. Using data obtained from a fixed bed microreactor, the product selectivity models were developed as functions of the above parameters via response surface methodology. The models were efficiently adjusted to avoid overfitting by considering cross-validation. The effects are shown via 3D diagrams. Single and multi-objective optimizations were then employed to maximize the production of DME and CO conversion and minimize the production of methanol, hydrocarbons, and carbon dioxide. Numerical optimization was performed through desirability charge ranging from zero to one where the highest desirability satisfies optimum conditions. Accordingly, an optimum area was obtained in which a variety of different points can be acceptable. Single-objective optimization provides a vaster area while multi-objective limits the feasible process conditions range.