Under excitation at 365 nm, the cell fluorescence is mainly due to bound and free NAD(P)H, plus a small contribution from flavins. Resolution is first attempted in the simplest case. i.e. the increase spectrum (AI,) due to microinjection of glucose-6-phosphate (G6P) into EL2 ascites cells. Above 510nm, AI, is identical to the spectrum of free NADH. Below 510nm. the presence of a second component is suggested, i.e. the intensity of the free NADH spectrum is lower than the measured AI, level. The difference between AI, and the free NADH spectrum (maximum at 475 nm) yields a spectrum suggestive of bound NADH with maximum at 450 nm. Thus, with free and bound NADH, the entire AI, can be reconstructed, with some assumptions as to the relative quantum yields of the two components. This seems to leave no place for a flavin component.The questions raised by the lack of such a component are answered using a new microspectrofluorometer, which allows correlated monitoring of NAD(P)H and flavins with excitations at 365 and 436 nm, respectively. As detected by excitation at 436 nm, injections of G6P, malate, ADP, and treatments with azide, cyanide O i partial anaerobiosis, all indeed show a redox change of flavins, in the sense of decreased emission. It is understandable, however, that such a change which is not very large even using 436 nm excitation should remain undetected when flavins are excited at 365 nm, i.e. using the tail of their excitation spectrum.In contrast to the increased AI, spectrum recorded in response to injected substrate, the initial spectrum ( I , ) of the cell prior to a metabolic perturbation reveals a third component, even with 365 nm excitation. The position and reactivity of this component shows flavin-like properties. The structural resolution attainable makes it possible to obtain the evaluation of free vs. bound NAD(P)H and flavin fluorochromes in the mitochondria1 and cytosolic compartments of the intact cell.