When cardiac muscle contracts and generates stress, heat is liberated. The heat-stress relation reveals two key indices relating to muscle performance. The intercept reflects the metabolic energy for removal of calcium from the cytosol and restoration of membrane potential, and hence is termed ‘activation heat’. The inverse of the slope of the heat-stress relation represents the energetic cost associated with each unit of stress development, and hence is termed ‘economy’. Despite over three decades of investigation, the heat-stress relation has never previously been reported at body temperature. Previous experiments have been performed below body temperature (typically 27 °C). In consequence, the heat-stress relation at 37 °C is still unknown. We aim to determine, for the first time, the heat-stress relation at 37 °C, and to compare it to that at 27 °C. We designed and constructed a calorimeter (Johnston 2014), the principle of which is based on the differential temperature of solution superfusing isolated cardiac tissue. We used thermoelectric sensors (Johnston et al. 2014) to achieve a 5-fold increase in thermal resolution. To measure the heat-stress relation, three‑month old male Wistar rats were deeply anaesthetised with isofluorane (5 % in oxygen, administered via inhalation according to protocol approved by the University of Auckland Animal Ethics Committee). Trabeculae were dissected from the right ventricle following cardiectomy, mounted in the calorimeter and stimulated at 3 Hz. Once muscle stress reached steady state, the entire calorimeter system was enclosed in a light-proof, temperature-controlled, enclosure. Active muscle stress development was varied by changing stimulus frequency and muscle length. Heat was simultaneously recorded. Heat was linearly correlated with stress for each of the six trabeculae studied. The heat‑stress regression lines were averaged, and the effect of temperature was examined, using the ‘random coefficient’ model (the regression coefficients of which are derived from ANCOVA using the statistical software package SAS). Values are means ± SEM, compared by ANCOVA. The heat-stress relation shifted down at the higher temperature. The change in slope was not significant (0.54 ± 0.07 to 0.60 ± 0.07), indicating no difference in the ‘economy’, but the ‘activation heat’ per twitch decreased from 3.5 kJ.m-3 ± 0.3 kJ.m-3 to 2.3 kJ.m-3 ± 0.3 kJ.m-3. These results reveal that the energetics of Ca2+ cycling is temperature-dependent, but that of crossbridge cycling in producing a unit of stress is not. In summary, we have made the first determination of the heat-stress relation of cardiac muscle at body temperature. The ability of our novel calorimeter to resolve tiny differences of temperature change at 37 °C promises better understanding of the energetics of cardiac muscle in health and disease.