The effect of the oxygen concentration on the rate of oxygen consumption by 786 and TA3 ascites tumor cell lines has been determined under steady-flow conditions with a membraneless fastresponding O2 electrode and using ascorbate and N,N,N',N'-tetramethyl-p-phenylenediamine as electron donors. The reaction was initiated by rapid injection of O2 into anaerobically incubated test system. The time-dependence of the intact cell respiration showed three distinct phases; an early very fast but short duration phase, a subsequent slow phase that prevailed for most of the reaction period and a third phase which preceded the reestablishment of anaerobiosis. Kinetic analysis of the reaction indicated a linkage between the catalytic efficiency and the transmembrane electrochemical potential. The rates of O2 uptake, obtained in the presence of both protonophores and ionophores, were monotonic and pseudo-first order over 90% of the course of O2 consumption. Extrapolation of the observed rates to zero time, at which zero A p H + and thus constant flow prevails, was used to calculate the oxygen concentration for the half-maximal respiratory rate, which was found to be in the range 1.55 -2.10 pM 02. No noticeable variation in the value of this kinetic parameter was found between the two cell lines used. Possible reasons for discrepancies in published reports on the oxygen dependence of the cytochrome c oxidase activity in various mitochondrial and reconstituted systems are discussed.ATP is required by living organisms to provide energy for biosynthesis, transport and mechanical work. It would seem that all tumor cells do not obtain the same proportion of ATP from mitochondrial and glycolytic reactions. The slowgrowth-rate and intermediate-growth-rate classes of tumor cells, as well as many normal tissues, obtain the bulk of their ATP from mitochondrial oxidative phosphorylation (more than 90% of the total), whereas rapidly growing tumors, even in the presence of a high glucose concentration, can obtain noticeable amounts of ATP from both sources, but with the higher proportion of ATP still being derived from oxidative phosphorylation (over 50%) (Pedersen, 1978). Moreover, accumulated evidence indicate that oxidation of glutamine is the major energy source for tumor cells, even in the presence of physiological levels of glucose (Moreadith andLehninger, 1984a, 1984b;Matsuno, 1987; Kovacevic et al., 1991). However, hexokinase bound to the mitochondria of highly glycolytic tumor cells utilises mitochondrially generated ATP rather than cytosolic ATP (Arora and Pedersen, 1988). Consequently, mitochondria from tumor cells would be the most important site for ATP synthesis.It is well established that a significant proportion of the cells in solid tumors of both rodents and human are hypoxic. Hypoxia has been identified as a factor that may affect the behavior of cancer cells, such as to limit the cure rate of standard radiotherapy and the action of some anticancer drugs in at least some types of human malignancy. Reoxygen-