Today, nonviral gene transfer vectors attract more attention as a therapeutic strategy for human diseases, because viral vectors such as adenoviral and herpes viral vectors have been proven to have problems, especially in immunogenicity and cytotoxicity. However, the main limitation of nonviral vectors has been low efficiency of gene expression. To overcome this defect, we have developed a new class of transfection vehicles, HVJ-cationic liposomes. The use of the cationic lipid DC-cholesterol facilitates efficient entrapment of negatively charged macromolecules (plasmid DNA, oligodeoxynucleotides, and proteins) and efficient interaction with negatively charged plasma membranes of cultured cells in vitro. Moreover, the fusogenic envelope proteins of hemagglutinating virus of Japan (HVJ) enhance delivery of the enclosed materials into cells. Using firefly luciferase as a marker, we optimized the liposome formula. As a result, we have succeeded in obtaining 100-800 times higher gene expression in vitro than with the conventional HVJ-anionic liposomes. However, in vivo gene transfer experiments have revealed that the use of cationic lipid instead of anionic lipid reduced the transgene expression dramatically in organs such as muscle and liver. We further discovered that the use of anionic liposomes with a viral-mimic king lipid composition increased transfection efficiency by several times in vivo. We conclude that the alternative usage of transfer vectors, for example, HVJ-anionic liposomes for in vivo delivery to extended areas of organs and HVJ-cationic liposomes for in vitro delivery (and possibly for in vivo delivery to a restricted area of organs), is of significance.