Improved bioluminescence analysis of pyridine nucleotides has been designed based on the fact that the luminescence intensity expresses the velocity of the light formation. The bacterial luciferase system is, in principle, composed of two reactions with two different velocities, one for energy supply by the oxidation of NAD(P)H and the other for the subsequent light generation. The rate setting can be arranged such that an emission maximum is produced 30 to 40 s after mixing the sample with the light yielding solution, hence providing for a convenient analytical performance. The maximal intensity which is easily recorded. e.g.. by a tracking volt-meter, is proportional to the concentration of the reduced nucleotide. Discriminative analysis of the various pyridine nucleotides is facilitated by selective destruction of the oxidized forms with alkali and the reduced forms with acid. Erroneous conversion of NAD(P)H to NAD(P)+ may be induced by haemoglobin in a tissue sample but this is prevented by the presence of 2 mM ascorbic acid at the instant of the acidification. Simultaneous coupling of the ongoing reduction of a pyridine nucleotide to the oxidation in the bacterial luciferase system generates a light-yielding cycle which offers important advantages. With NAD(P)+ as the analytic target compound. direct measurement replaces a preceding separate conversion to NAD(P)H. The four nucleotide forms become determinable in a sample by combining selective destruction of either the reduced or oxidized species with a nucleotide-specific reduction in the cycle. Discriminative analyses are furthermore facilitated by the enhanced emission which is due to the energy derived from the continuous specific reduction, whereas initial light signals from side reactions fade out. It is often possible to suppress disturbing analytical errors by the design of the light-yielding cycle. If the rate of the dehydrogenase reactions is kept low compared with the overall rate of the luciferase system, moderately impaired function of some of its components may only give rise to a slight and tolerable decrease in emission intensity. Kinetic evaluations and model experiments are presented and supplemented with applications to tissue samples.