A kinetic method for the determination of O2 solubility in air-saturated aqueous solutions of widely varying composition and temperature is described. It is based on the precise molar stoichiometry between the rates of uptake of H+ and O2, measured with response-matched electrodes, in the reaction NADH + H+ + 1/2O2----NAD+ + H2O, catalyzed by an NADH oxidase preparation. To the initially anaerobic test system, which contains an excess of NADH and NADH oxidase in a buffered medium, an aliquot of the O2-containing solution to be tested is added and the rates of both O2 uptake and H+ uptake are recorded; the H+ electrode is calibrated against standard HCl. From these data the amount of O2 in the aliquot is calculated. Some representative values for O2 solubility at 25 degrees C and 760 mm in air-saturated systems are (i) distilled H2O, 516 nmol O/ml, (ii) 0.15 M KCl, 480 nmol O/ml, and (iii) 0.25 M sucrose, 458 nmol O/ml. Data and equations are also given for the solubility of O2 at 760 mm in air-saturated and lightly buffered 0.15 M KCl and 0.25 M sucrose over the range 5 to 40 degrees C. In the method described the rates of O2 and H+ uptake are precisely linear and stoichiometric when NADH is present in large excess over O2. However, when O2 is in excess and small additions of 340-nm-standardized NADH are made, as in earlier methods based on NADH oxidation, the endpoint is approached very gradually and tends to overestimate O2 solubility, owing to (i) the higher Km for NADH than for O2, (ii) the relatively slow response of the Clark O2 electrode, and (iii) the incomplete oxidation of NADH in the presence of 340-nm-absorbing inhibitory substances.