Modifying cell-penetrating peptides (CPPs) with fatty acids has long been used to improve peptide-mediated nucleic acid delivery. In this study we have revisited this phenomenon with a systematic approach where we developed a structure-activity relationship to describe the role of the acyl chain length in the transfection process. For that we took a well-studied CPP, PepFect14, as the basis and varied its N-terminal acyl chain length from 2 to 22 carbons. To evaluate the delivery efficiency, the peptides were noncovalently complexed with a splice-correcting oligonucleotide (SCO) and tested in HeLa pLuc705 reporter cell line. Our results demonstrate that biological splice-correction activity emerges from acyl chain of 12 carbons and increases linearly with each additional carbon. To assess the underlying factors regarding how the transfection efficacy of these complexes is dependent on hydrophobicity, we used an array of different methods. For the functionally active peptides (C12-22) there was no apparent difference in their physicochemical properties, including complex formation efficiency, hydrodynamic size, and zeta potential. Moreover, membrane activity studies with peptides and their complexes with SCOs confirmed that the toxicity of the complexes at higher molar ratios is mainly caused by the free fraction of the peptide which is not incorporated into the peptide/oligonucleotide complexes. Finally, we show that the increase in splice-correcting activity correlates with the ability of the complexes to associate with the cells. Collectively these studies lay the ground work for how to design highly efficient CPPs and how to optimize their oligonucleotide complexes for lowest toxicity without losing efficiency.