Polyamide nucleic acids (PNAs) have emerged as useful agents for recognition of single- and double-stranded nucleic acids. Interresidue hydrogen bonds between the amide carbonyl nearest the nucleobase and chain NH moieties provide inherent stability to the helical conformation of PNA 1. Moving the amide carbonyl away from the nucleobase to the backbone, and replacing it with a methylene group, results in 2 lacking the stabilizing hydrogen bond. Oligomers of 2 do not interact with DNA. Modeling suggests that 2 displays a more extended conformation than 1, and nucleobase orientation is disrupted in 2 in the absence of a complementary DNA strand. This is in contrast to 1, which retains a centrosymmetric arrangement of nucleobases. Structures for 1-T10.DNA and (1-T10)2.DNA species spanned by a pyrimidine strand (D-loop) were constructed. In the triple helical (1-T10)2.DNA structure, the two PNA strands form the complementary Watson-Crick paired strand and the Hoogsteen base-paired strand in the major groove of the 1.DNA duplex. The PNA strands are proposed to bind antiparallel to one another in (1-T10)2.DNA structure. The factors suggested to account for the stability of this 2:1 complex are (i) a hydrophobic attraction between two PNA backbones and (ii) a favorable electrostatic effect resulting from replacement of a phosphodiester backbone by a neutral peptide backbone.