Abstract Potentiometric, spectrophotometric and kinetic techniques have been used to determine the stabilities of complexes of transition metal ions with phosphoric acid and ribose-phosphate, purine nucleosides and nucleotides [1–4]. At ionic strenght 0.1 M, dinegative phosphate groups bind to Ni 2+ and Co 2+ with stability constants close to 100 M −1 [1, 5]. The neutral purine nucleosides form only weak complexes; the binding constants depend markedly on the base involved, e.g. K = 14 for Ni 2+-inosine, and K = 2 M −1 for Ni 2+-adenosine . The data are consistent with assumption that the N7 atom of the imidazole ring is the predominant binding site. Similar differences are observed also for the complex stabilities of the nucleotides: K(NiIMP) = 920 M −1, and K(Niz.sbnd;AMP)=300 . The experimental overall stabilities of the nucleotide complexes can be rationalized only by assuming a chelate structure, with the metal ion being to the phosphate group and to the base. Space-filling models indicate that in the nucleotide complexes only the N7 atom an act as the binding site of the base. The kinetic data, too, can be interpreted only by a stepwise chelate formation process. Moreover, the kinetic data enable also the evaluation of the stepwise equilibria. In the case of NiAMP, about 2 3 of the complexes are present in the chelate form, 1 3 in the monodentate form. Complex formation of Ni 2+ with the dinucleoside-monophosphate ApA − is weaker (K = 2.6 M −1) than with AMPH − (K = 11) , despite the availability of an additional adenosine group in ApA −. This observation is attributed to the conformational properties of ApA − in solution. the ligand prefers a conformation in which the two adenines are in a stacked position. In this conformation the N7 atoms are far away from the phosphate group, and the metal ion can bind either to the phosphate or to the base, but not simultaneously to both . Strong interactions were observed between Ni 2+ ions and the polynucleotide poly(A) . Quantitative data were obtained by using murexide as an indicator for the concentration of free metal ions. The difference from the total metal concentration gives the amount of Ni 2+ bound to poly(A). This technique is applicable for free Ni 2+ concentrations 10 −5–10 −3 M. At the upper limit, 1 Ni 2+ is bound per 4.4 mononucleotide units (ionic strenght 0.1 M). A Scatchard plot yields a straight line, i.e. the process can formally be described as the binding of Ni 2+ to one class of independent binding sites at poly(A). Slope and intercept of the plot yield a value of 0.26 for the number of binding sites per monomer ( i.e. 4 monomers form one binding site), and a stability constants K = 8.2 × 10 3 M −1 for the binding of Ni 2+ to these binding sites. In kinetic studies of this system two reaction effects were detected. The kinetic data can be rationalized in terms of a mechanism which involves a second-order (outer-sphere) association process followed by 3 first-order steps (inner-sphere binding, probably to two phosphates and one base ). The high overall stability constant is mainly due to strong outer-sphere association (high charge density at poly(A)).