PURPOSE: We tried to assess the optimal conditions to improve low transduction efficiency and their effect on target cells. METHODS: Cultured NIH 3T3 cells were incubated with retroviral vectors bearing an enhanced green fluorescent protein (eGFP) gene. We varied the ratio of viral vectors to target cells (1:1-1:8) and the number of transfections (x1, x2), and compared transduction efficiencies. Also, the effects of polybrene on transduction efficiency and viability of target cells were assessed. Transduction of the eGFP gene was evaluated by observing NIH 3T3 cells under a fluorescence microscope and efficiencies were measured by the percentage of eGFP positive cells using FACscan. RESULTS: As the ratio of retroviral vectors to target cells increased, transduction efficiency was greatly improved, from 7% (1:1) to 38% (1:4). However, transduction efficiency did not increase any more when the ratio increased from 1:4 to 1:8. Cells transfected twice showed higher transduction efficiencies than cells transfected once, at a ratio of 1:8. The eGFP gene transduced to NIH 3T3 cells sustained its expression during repeated passages. However, after the third passage (day 9), the percentage of eGFP positive cells began to decline. The degree of this decline in eGFP expression was lower in cells transfected twice than in cells transfected once (P<0.05). The addition of polybrene did not have any toxic effect on NIH 3T3 cells and greatly increased transduction efficiency (P=0.007). In addition to vector component, transduction efficiency was very sensitive to culture confluence. Cells cultured and transfected in 24-well plate showed higher transduction efficiency, although cells cultured in 6- well plate proliferated more (P=0.024). CONCLUSION: Our data could be used as a basis for retrovirus-based gene therapy. Further study will follow using human cells as target cells.