A previous study has shown an efficient, systemic transgene expression in mice via intravenous administration of a LPD formulation composed of DOTAP liposomes, protamine sulfate and plasmid DNA. In this study, factors affecting the in vivo performance of this formulation were further evaluated. A protocol in which liposomes were mixed with protamine before the addition of plasmid DNA was shown to produce small condensed particles with a diameter of about 135 nm. These particles were stable over time and gave a high level of gene expression in all tissues examined including lung, heart, spleen, liver and kidney with the highest level of expression in the lung. Inclusion of dioleoylphosphatidylethanolamine (DOPE) as a helper lipid significantly decreased the in vivo activity of LPD. In contrast, inclusion of cholesterol as a helper lipid increased the in vivo transfection efficiency of LPD and more importantly, decrease the amount of cationic lipid required for the maximal level of gene expression. Studies on the interaction between mouse serum and LPD showed that LPD became negatively charged after exposure to serum, and LPDs containing different helper lipids varied in the amount of associated serum proteins. LPD containing DOPE was more enriched in a protein corresponding to albumin in molecular weight. These results suggest that the mechanism of LPD-mediated intravenous gene delivery might be different from that of in vitro lipofection and that serum protein association might be a major factor limiting the in vivo transfection by LPD.