The transcription factor “decoy” (TFD) strategy has been demonstrated as one of the ways to artificially modulate gene expression at the transcription level. This approach is based on the competition for transcription factors between endogenous cis-elements present within regulatory regions of target genes and exogenously added DNA sequences (the DNA-based drug) mimicking the specific cis-elements. With respect to pharmacogenetic therapy based on DNA and DNA mimics, peptide nucleic acids (PNAs) are of great interest. In these molecules, the sugar-phosphate backbone of DNA is replaced with a pseudopeptide chain. PNAs hybridise to DNA and RNA with high efficiency forming highly stable PNA/DNA and PNA/RNA duplexes and are resistant to proteinases and exo- and endonucleases. For these reasons, they are proposed as very important tools in gene therapy. Despite the fact that the antigéne as well as antisense potential of therapeutic PNA and PNA analogues is well known, the possible use of PNA-based molecules in TFD experiments has been only marginally studied. This is mainly due to the altered structure of the DNA/PNA and PNA/PNA hybrids, leading to great differences in the major groove and affecting direct binding of transcription factors to target PNA-based molecules, as well as stability of the generated complexes. With respect to the TFD approach, promising results have been on the contrary obtained using double-stranded PNA-DNA chimeras. PNA-DNA-PNA chimeras mimicking NF-kB and Sp1 binding sites present within the long terminal repeat of the human immunodeficiency type 1 virus were tested for decoy activity. The results obtained show that these molecules are more soluble than PNAs, exhibit efficient decoy activity, are more resistant than DNA/DNA hybrids to nucleases and suitable for delivery with cationic liposomes and microspheres.