We have shown that the rate of NADH-coenzyme Q reductase in rat liver mitochondria, assayed using the decyl-ubiquinone analog DB, is underevaluated, probably as a result of its low water solubility. In view of drawbacks encountered using other more soluble acceptors in this system, we demonstrate that the most reliable assay of the physiological rate of CoQ reduction by Complex I is the indirect calculation from the total rate of NADH oxidation and the rate of ubiquinol oxidation, using the pool equation of Kröger and Klingenberg [(1973) Eur. J. Biochem. 34, 358-368].