Although it has been known for many years that isolated respiring mitochondria have a massive capacity to accumulate calcium, the physiological significance of this pathway has remained uncertain until very recently. In the last few years, it has become clear that, in most cells, physiological Ca²⁺ signals are associated with significant mitochondrial Ca²⁺ uptake. Under pathological conditions, mitochondrial Ca²⁺ uptake may be a crucial step in the progression to cell death. During episodes of hypoxia in the CNS, the excitatory transmitter glutamate accumulates in the extracellular space. Prolonged stimulation of glutamate (and particularly NMDA) receptors is toxic to neurons, mediated largely by an increase in intracellular calcium concentration ([Ca²⁺]_c). We have made simultaneous measurements of mitochondrial potential (Δσ_m) and intracellular [Ca²⁺]_c in rat hippocampal neurons in culture. Application of glutamate sufficient to cause 70Ð80 % cell death was associated with the collapse of Δσ_m. In contrast, raising [Ca²⁺]_c simply through depolarisation with 50 mM KCl had no significant impact on Δσ_m, even when the [Ca²⁺]_c levels achieved were similar to those seen with glutamate. Examination of the glutamate responses cell by cell showed a highly heterogeneous mitochondrial response, both in time course and amplitude while the [Ca²⁺]_c responses were highly homogeneous. Thus, the mitochondrial responses did not show a simple quantitative relationship to the change in [Ca²⁺]_c. We therefore sought other variables that might contribute to the response. Obvious candidates were superoxide and NO. No increase in superoxide production (measured with hyodroethidine fluorescence) was detectable in response to glutamate and the mitochondrial response to glutamate was not altered by a range of antioxidant scavengers (Vergun et al. 2001). Notably, a combination of antioxidants inhibited the glutamate current and suppressed the glutamate-induced rise in [Ca²⁺]_c, suggesting that great care must be used in interpreting data using such reagents. Inhibition of nNOS suppressed the change in Δσ_m (Keelan et al. 1999) without altering the [Ca²⁺]_c response. Further, in the presence of an NO donor, the [Ca²⁺]_c response to 50 mM KCl now caused a mitochondrial depolarisation. Thus it seems that the collapse of Δσ_m with glutamate requires the synergistic action of both NO and Ca²⁺. The ‘source specificity’ of glutamate-mediated increase in [Ca²⁺]_c compared with that mediated by KCl is probably mediated by the localisation of nNOS close to NMDA receptors by the scaffolding protein PSD-85 (Sattler et al. 1999) so that domains of [Ca²⁺]_c sufficient to activate nNOS only occur close to NMDA receptors and not close to voltage-gated Ca²⁺ channels. The other major issue to consider is the mechanism underlying the collapse of mitochondrial potential. One likely candidate is the opening of the mitochondrial permeability transition pore (mPTP), a channel which opens in the mitochondrial inner membrane under pathological conditions of high [Ca²⁺]_m, oxidative stress, high [P_i] and ATP depletion. mPTP opening is blocked by cyclosporin A (CsA) which inhibited the glutamate-induced collapse of mitochondrial potential. However, CsA also inhibits calcineurin, so inhibiting NO production, and the mitochondrial depolarisation was also suppressed by FK506, which inhibits calcineurin but does not affect the mPTP. It therefore remains difficult to attribute the glutamate-induced loss of mitochondrial potential unequivocally to mPTP opening without more precise experimental tools.