This paper reviews the model of the control of mitochondrial substrate oxidation by Ca2+ ions. The mechanism is the activation by Ca2+ of four mitochondrial dehydrogenases, viz. glycerol 3-phosphate dehydrogenase, the pyruvate dehydrogenase multienzyme complex (PDH), NAD-linked isocitrate dehydrogenase (NAD-IDH) and 2-oxoglutarate dehydrogenase (OGDH). This results in the increase, or near-maintenance, of mitochondrial NADH/NAD ratios in the activated state, depending upon the tissue and the degree of 'downstream' activation by Ca2+, likely at the level of the F1Fo ATPase. Higher values of the redox span of the respiratory chain allow for greatly increased fluxes through oxidative phosphorylation with a minimal drop in protonmotive force and phosphorylation potential. As PDH, NAD-IDH and OGDH are all located within the inner mitochondrial membrane, it is changes in matrix free Ca2+ [Ca2+]m which act as a signal to these activities. In this article, we review recent work in which [Ca2+]m is measured in cells and tissues, using different techniques, with special emphasis on the question of the degree of damping of [Ca2+]m relative to changes in cytosol free Ca2+ in cells with rapid transients in cytosol Ca2+, e.g. cardiac myocytes. Further, we put forward the point of view that the failure of mitochondrial energy transduction to keep pace with cellular energy needs in some forms of heart failure may involve a failure of [Ca2+]m to be raised adequately to allow the activation of the dehydrogenases. We present new data to show that this is so in cardiac myocytes isolated from animals suffering from chronic, streptozocin-induced diabetes. This raises the possibility of therapy based upon partial inhibition of mitochondrial Ca2+ efflux pathways, thereby raising [Ca2+]m at a given, time-average value of cytosol free Ca+2.