Changes in neuronal energy metabolism, mitochondrial functions and homeostasis of reactive oxygen species are often supposed to induce alterations in neuronal activity in hippocampal slice models. In order to investigate the NAD(P)H autofluorescence signal in brain slice models, methods to monitor NAD(P)H signal in isolated mitochondria as described by Chance et al. [J. Biol. Chem. 254 (1979) 4764] and dissociated neurons as described by Duchen [Biochem. J. 283 (1992) 41] were adapted to recording conditions required for brain slices. Considering different experimental questions, we established an approach to monitor NAD(P)H autofluorescence signals from hippocampal slices of 400 microm thickness under either submerged or interface conditions. Therefore the procedure described here allows the measurement of NAD(P)H autofluorescence under conditions typically required in electrophysiological experiments. Depolarization of plasma membrane caused by electrical stimulation or application of glutamate (100 microM) resulted in a characteristic initial decrease followed by a long-lasting increase in the NAD(P)H autofluorescence signal. H(2)O(2) (100 microM) evoked a strong NAD(P)H signal decrease indicating direct oxidation to the nonfluorescencend NAD(P)(+). In contrast, the increase in NAD(P)H signal that followed a brief inhibition of mitochondrial respiratory chain complex I using rotenone (1 microM) indicated an accumulation of NAD(P)H. However, in presence of rotenone (1 microM) electrically evoked long-lasting NAD(P)H signal overshoot decreased progressively, due to a negative feedback of accumulated NAD(P)H to the citrate cycle. A comparable reduction in NAD(P)H signal increase were observed during low-Mg(2+) induced epileptiform activity, indicating a relative energy failure. In conclusion, the method presented here allows to monitor NAD(P)H autofluorescence signals to gain insight into the coupling of neuronal activity, energy metabolism and mitochondrial function in brain slice models.