Insect Drosophila melanogaster S2 cells were developed as a plasmid-based and therefore nonlytic expression system for functional foreign proteins. Transfection is an important step to introduce foreign target DNA into cells and should be properly optimized to obtain maximum production yield. Single factor search (SFS) methodology is still generally used to determine optimal condition in a biological system. Although this method is relatively simple to perform it has many disadvantages such as not considering interactions between several factors and not covering the entire region of the solution pool. Therefore, we approached this optimization problem statistically with response surface (RSM) and evolutionary operation (EVOP) methodologies and compared the transfection efficiencies with the traditional SFS method. We employed secreted green fluorescent protein (GFP) as a reporter for determination of optimal transfection condition and secreted human erythropoietin (hEPO) as a confirming foreign model protein. Consequently, we arrived at the best optimal transient transfection condition (1 mug of plasmid DNA, 5 mug of lipofectin, 2 x 10(6) cells of initial cell number, and 18 h of transfection duration time) through a systematic access in a series of SFS, RSM, and EVOP. The secreted hEPO yield using optimal transient transfection condition by EVOP methodology was enhanced by about 1.8-fold compared to that of traditional SFS. This optimized transient transfection condition can be used as a basis for optimal stable transfections. A linear relationship between secreted GFP fluorescence intensity and secreted hEPO concentration indicated that facile and noninvasive determination of optimal transfection. conditions for expression and secretion of foreign proteins in S2 cell cultures was made possible by simple measurement of GFP fluorescence.