Although adenoviral vectors (Ad) have proven effective for gene therapy in animal models of cardiovascular diseases, the ability to achieve comparable gene transfer with nonviral vectors would provide potentially desirable safety and toxicity features for clinical studies. In this report, we optimized the new cationic liposome (CL) GAP DLRIE (GDD) for gene transfer in injured pig arteries and compared its efficiency to DMRIE/DOPE (DD), a standard CL, and DNA alone, using a plasmid encoding a chloramphenicol acetyltransferase gene (CAT). Optimal GDD/DNA formulation was compared to a recombinant Ad encoding CAT (Ad-CAT). A plasmid encoding the human alkaline phosphatase (hAP) gene was used, in the optimized conditions, to localize gene transfer. Gene expression was assessed 48 hours following delivery. Dose response studies were performed by varying the ratio and quantity of GDD (31.6 to 947 μM) and DNA (20.15 to 1210 μM). The optimal GDD to DNA ratio (0.78 molar ratio) corresponds to an excess of positively charged CL to negatively charged DNA, due to the presence of 2 positive charges on each GDD molecule. At the 0.78 ratio, a maximum CAT activity was reached with 631 μM GDD/806 μM DNA. In these optimized conditions, hAP gene expression was observed in initmal cells (6%) and medial smooth muscle cells (<1%). Optimal GDD/DNA formulation provided a 15-fold higher level of gene expression compared to DNA alone and DD. Compared to gene transfer obtained with the highest titers (10 1 0 pfu/ml) of Ad-CAT, the GDD/DNA formulation resulted in 20 fold less gene expression. Improvements in arterial gene expression can be achieved by optimization of transfection conditions with CL- DNA complexes in vivo which may prove useful for gene delivery in cardiovascular diseases.