Targeted gene transfer, addressing the alphavbeta3 integrin by coupling the appropriate ligand, cRGD (S(2)-bridged cyclic Arg-Gly-Asp containing peptide) motif, on to a DNA condensing sequence was described as early as 1995 by Hart, Harbottle, Cooper, Miller, Williamson and Coutelle [(1995) Gene Ther. 2, 552-554]. Their work was followed by a series of publications, introducing the cRGD motif in polycationic DNA carriers, such as peptides, proteins and liposomes. Polyethylenimine and even adenoviruses were additionally ligated using the cRGD motif. 'Integrin specificity' has been determined from the significantly improved transfection efficiency compared with the DNA carriers with control ligands, mainly the cRGE (S(2)-bridged cyclic-Arg-Gly-Glu-containing peptide) motif. However, by observing the physicochemical appearance of the resulting complexes and their controls such as the poly(L-lysine)-DNA complexes carrying the cRGD and the cRGE motifs, we doubted the integrin-mediated specificity of the increased transfection efficiency. To clarify this contradiction, we investigated the suitability of the cRGD motif for targeted gene transfer. We proved the specificity of the RGD motif and its controls using computational docking procedures and molecular modelling methods. Since we were confident of the motifs used, we improved our transfection method. Since aggregation of the RGD-ligated poly(L-lysine)-DNA complexes under physiological conditions caused an enormous amount of unspecific cell uptake and transfection, a method had to be designed to exclude aggregation processes of the motif-polycation-DNA complexes. Small complex sizes are necessary for receptor-specific uptake. The complexes were therefore recharged using poly(vinyl sulphate). Inhibited aggregation of the targeted DNA carriers under physiological conditions is a necessary prerequisite for successful in vivo gene transfer.