Lipoplex nano-aggregates constituted of plasmid DNA (pDNA) pEGFP-C3 and mixed cationic liposomes, consisting of several percentages of a gemini cationic lipid (GCL) of the 1,2-bis(hexadecyl imidazolium) oxyethylene series, referred to as (C(16)Im)(2)(C2O)(n), with oxyethylene spacers (n = 1, 2 or 3) between the imidazolium cationic groups and the DOPE zwitterionic helper lipid, have been characterized by various biophysical and biological approaches carried out at several GCL compositions (alpha), and either the mass or the effective charge ratio of the lipoplex. The electrochemical study by zeta-potential confirms that the three GCLs yield a 10% lower effective charge than the nominal one, while compacted pDNA yields only a 25% effective negative charge. The SAXS study reveals, irrespective of the spacer length (n) and effective charge ratio (rho(eff)), the presence of two lamellar structures, i.e., one (L-alpha,L-main) in the whole GCL composition and another (L-alpha,L-DOPE,L-rich) with higher periodicity values that coexists with the previous one at low GCL composition (alpha = 0.2). The cryo-TEM analysis shows two types of multilamellar structures consisting of cationic lipidic bilayers with pDNA sandwiched between them: a cluster-type (C-type) at low alpha = 0.2 and a fingerprint-type (FP-type) at alpha >= 0.5, both with similar interlamellar spacing (d) in agreement with the L-alpha,L-main structure determined by SAXS. Transfection efficacies (TEs) of each lipid mixture were determined in four different cell lines (HEK293T, HeLa, Caco-2 and A549) at several alpha and rho(eff) values in the absence and presence of serum (FBS). The optimized formulations (alpha = 0.2 and rho(eff) = 2.0) substantially transfect cells much better than a commercial transfection reagent, Lipofectamine 2000 and previously studied efficient lipoplexes containing other cationic head groups or spacers both in the absence and presence of serum. The activity of optimized formulations may be attributed to the combination of several factors, such as: (a) the fusogenic character of DOPE which results in higher fluidity of the lipoplexes at alpha = 0.2, (b) the coexistence of two lamellar structures at alpha = 0.2 that synergizes the TE of these lipid vectors, and mainly (c) the higher biocompatibility of the GCLs reported in this work due to the presence of two imidazolium cationic groups together with an oligo-oxyethylene spacer. The length of the spacer in the GCL seems to have less impact, although (C(16)Im)(2)(C2O)(n)/DOPEpDNA lipoplexes with n = 1 and 3 show higher gene transfection than n = 2. All the optimum formulations reported herein are all highly efficient with negligible levels of toxicity, and thus, may be considered as very promising gene vectors for in vivo applications.