DNA delivery systems for gene therapy applications have to be able to trigger the uptake of plasmid DNA into the nucleus. We have tested two types of non-viral vector systems, lipofection (cationic lipid-based, using Lipofectamine) and polyfection (cationic polymer-based, using glycerol enhanced transferrinfection), for their ability to transfect confluent, contact inhibited primary human fibroblasts. While both systems worked well with growing fibroblasts, polyfection was superior with confluent cells. A slight reduction in cell associated plasmid DNA was observed with resting cells, but it was similar for both types of complexes. Lipofectamine showed a prevalence for transfecting cycling cells as judged by costaining transfected cells with cell cycle markers. No such bias was observed when glycerol enhanced transferrinfection was used. Microinjection of plasmid DNA/polylysine complexes into the cytoplasm of fibroblasts resulted in a higher percentage of expressing cells than injection of plasmid DNA, offering an explanation for the higher transfection levels obtained with transferrinfection in non-growing cells.