In this study, we compared two systems which can be applied for transfection in vitro and in vivo: polyplexes based on the polymer dextran-spermine (D-SPM) and lipoplexes based on 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP)/cholesterol. The carriers differ in (1) solubility in aqueous media, (2) source of the positive charges (quaternary amines for DOTAP and primary plus secondary amines for D-SPM), (3) electrostatics, i.e., for lipid and polymer, respectively: zeta-potential (81.0 and 48.1 mV), surface potential (180 and 92 mV), and surface pH (10.47 and 8.97), and (4) charge distribution (ordered versus non-ordered). The stability of the complex upon interaction with serum proteins was studied by means of fluorescence resonance energy transfer (FRET) between rhodamine-labeled cationic carriers and fluorescein-labeled DNA. Addition of serum increases the lipid-DNA average distance and decreases the polymer-DNA distance. However, FRET efficiency indicates that serum proteins do not induce a major DNA dissociation for either polyplexes or lipoplexes. Comparing the biodistribution of rhodamine-labeled complexes and the transgene expression after intravenous (i.v.), intramuscular (i.m.), and intranasal (i.n.) administration, we found that local administration of lipoplexes resulted in the lipoplexes remaining at the site of injection, whereas the polyplexes showed systemic distribution, accompanied by transgene expression in lungs and liver. It is suggested that the high water-solubility of the polymer combined with its lower positive charge (compared to DOTAP), which makes its association with the cells at the site of injection weaker, enables the polymer to reach and transfect distant organs through the blood stream. Using chemically modified D-SPM, we demonstrated the importance of high density of positive charges and a sufficient level of secondary amines for achieving efficient transgene expression in vivo.