CD spectra of poly(dA-dT).poly(dA-dT) in low salt (10-100 mM NaCl) and high salt (4-6 M CsF) are different i.e. 275 nm band gets inverted in going from low to high salt (Vorhickova et. al., J. Mol. Biol. 166, 85, 1983). However, from CD spectra alone it is not possible to decipher any structural differences that might exist between the low and high salt forms of poly(dA-dT).poly(dA-dT). Hence, we took recourse to high resolution NMR spectroscopy to understand the structural properties of poly(dA-dT).poly(dA-dT) in low and high salt. A detailed analysis of shielding constants and extensive use of NOE studies under minimum spin diffusion conditions using C(8)-deuterated poly(dA-dT).poly(dA-dT) enabled us to come up with the following conclusions (i) base-pairing is Watson-Crick under low and high salt conditions. (ii) under both the conditions of salt the experimental data can be explained in terms of an equilibrium blend of right and left-handed B-DNA duplexes with the left-handed form 70% and the right-handed 30%. In a 400 base pairs long poly(dA-dT).poly(dA-dT) (as used in this study), equilibrium between right and left-handed helices can also mean the existence of both helical domains in the same molecule with fast interchange between these domains or/and unhindered motion/propagation of these domains along the helix axis. (iii) However, there are other structural differences between the low and high salt forms of poly(dA-dT).poly(dA-dT); under the low salt condition, right- and left-handed B-DNA duplexes have mononucleotide as a structural repeat while under the high salt conditions, right- and left-handed B-DNA duplexes have dinucleotide as a structural repeat. In the text we provide the listing of torsion angles for the low and high salt structural forms. (iv) Salt (CsF) induced structural transition in poly(dA-dT).poly(dA-dT) occurs without any breakage of Watson-Crick pairing. (v) The high salt form of poly(dA-dT).poly(dA-dT) is not the left-handed Z-helix. Although the results above from NMR data are quite unambiguous, a question still remains i.e. what does the salt (CsF) induced change in the CD spectra of poly(dA-dT).poly(dA-dT) really indicate? Interestingly, we could show that the salt (CsF) induced change in poly(dA-dT).poly(dA-dT) is quite similar to that caused by a basic polypeptide viz. poly-L(Lys2-Ala)n i.e. both the agents induced a psi-structure in DNA.