One of the major questions in any gene therapy approach is the selection of the appropriate vector system. Here, the optimization of a gene transfer protocol for renal cell carcinoma using lipofection as a nonviral gene transduction system was evaluated. To select the promoter which gives the highest expression, different plasmids which are able to express Escherichia coli beta-galactosidase gene as a reporter gene under the control of different promoters were tested: human cytomegalovirus promoter (pCMVbeta), simian virus 40 promoter (pSVbeta), adenovirus promoter (ADbeta), and herpes simplex virus thymidine kinase promoter (TKbeta). The pCMVbeta revealed the highest expression of the beta-gal gene in the renal cell carcinoma (RCC) lines. Thus this CMV promoter was selected for the expression of the granulocyte-macrophage colony stimulator factor (GM-CSF) gene. Three different lipids (LipofectAmine, LipofectAce, and Lipofectin) were compared for their transduction efficiency, and the optimal conditions for quantitatively high lipofection rates were established. The consistently best results regarding gene expression as well as viability of the RCC lines were obtained when Lipofectin was used. Gene expression was monitored by a specific enzyme-linked immunosorbent assay and functionally validated by a cell proliferation test. The GM-CSF expression profile showed a peak at 48 hours after transfection and was still detectable after 5 days. Here the feasibility of efficient lipofection of the GM-CSF gene into RCC lines is demonstrated. Most importantly, considerable differences in the relative quantity of GM-CSF gene transfer into the different RCC lines was observed here. This may be of critical relevance for the design of any clinical gene transduction protocol in tumor cell vaccination attempts.