Segments of isolated intact rat mesenteric small arteries were incubated in physiological bicarbonate buffer in the presence of nano- to millimolar concentrations of ATP. ATP was hydrolysed, and when the vessel was transferred from one incubation to another, the enzyme activity was transferred with the vessel, consistent with the presence of an ecto-ATPase. The substrate, ATP, was shown to induce a modification of the hydrolytic activity which occurred the more rapidly the higher the concentration of ATP. The modified system hydrolysed ATP with a decreased substrate affinity. As the substrate induced a modification of the hydrolytic activity, steady-state velocity measurements for determination of kinetic parameters could not be obtained. Nevertheless, it was possible to compare the modification caused by ATP and UTP, and to compare the hydrolysis rates measured with [ 32P]ATP, [ 32P]UTP and [ 32P]GTP. It was concluded that the hydrolytic activity of the vessels did not distinguish between the nucleoside triphosphates (NTPs). In a histidine buffer, the activity was shown to be activated by micromolar concentrations of either Ca 2+ or Mg 2+, and not to be influenced by inhibitors of P-type, F-type and V-type ATPases. Functional removal of the endothelium before assay did not reduce the measured NTP hydrolysis. At millimolar concentrations of trinucleotide the hydrolysis rate was 10–15 μmol per min per gram of tissue or 0.11-0.17 μmol per min per 10 6 vascular smooth muscle cells. This value is equivalent to the maximal velocity obtained for the Ca 2+ or Mg 2+-dependent NTPase released to the medium upon 2 s of sonication of the vessels (Plesner, L., Juul, B., Skriver, E. and Aalkaj˦r, C. (1991) Biochim. Biophys. Acta 1067, 191–200). Comparing the characteristics of the released NTPase to the characteristics of the activity of the intact vessel, they showed a strong resemblance, but the substrate-induced modification of the enzyme was seen only in the intact preparation.