Heating and cooling alter local limb tissue blood flow and its surrogate oxygen and blood supply in humans and animals, but the temperature sensitive mechanisms remain unknown. In the clinical and athletic settings heating [1] and cooling [2] are routinely used to improve recovery or alleviate symptoms. We have recently demonstrated that passive heating increases limb tissue blood flow in humans in association with the increases in muscle temperature and arterial plasma adenosine 5′-triphosphate (ATP) [3, 4] and that the erythrocytes are the sole blood source of ATP (3). However, the effect of cooling on erythrocyte ATP release has never been investigated. Here we tested the hypothesis that the release of the vasodilator and sympatholytic mediator ATP from human erythrocytes is sensitive to both increases and reductions in blood temperatures. To accomplish this aim, erythrocytes, plasma and serum were isolated from blood taken from 12 volunteers. To investigate the effects of heating and cooling, samples were exposed to physiological temperatures in water baths set at 39°C (hot), 33°C (control), 27°C (cool) and 20°C (coldest) for 20 mins. ATP was measured after the addition of a “stop solution” which prevented any further release or degradation and analysed by using a luciferin-luciferase technique while free haemoglobin was assessed as an indicator of haemolysis. Values reported are means ± SEM for n=12. Erythrocyte ATP release was increased from 0.61 ± 0.13 µmol/L at 33°C to 1.14 ± 0.19 µmol/L at 39 °C (P<0.05). ATP release from erythrocytes exposed to 27°C was 0.49 ± 0.08 µmol/L whilst at the coldest temperature of 20°C ATP release was significantly reduced to 0.39 ± 0.05 µmol/L (P<0.05) compared to control. Consequently, the release of ATP from erythrocytes was tightly associated with changes in temperature from 20°C to 39°C (r2 = 0.97; P<0.05). Free haemoglobin ranged from 0.079 ± 0.007 g/L at 33°C, 0.073 ± 0.012 g/L at 39°C, 0.074 ± 0.014 g/L at 27°C and 0.060 ± 0.007 g/L at 20°C indicating that temperature did not increase haemolysis. In contrast, no changes in ATP levels were observed in either isolated plasma or serum samples at these temperatures. In conclusion, these results demonstrate that erythrocyte ATP release is sensitive to physiological changes in temperature produced by local heating and cooling. Furthermore, they imply that heat and cooling can be used as a non-pharmacological means of manipulating the supply of oxygen and blood to human tissues for therapeutic benefit.