Occupational workers and military personnel who perform their activities in hot environmental conditions are not only at risk of hyperthermia, but the cardiovascular adjustments that occur in such settings can also place them at risk for cardiovascular instability, particularly should they suffer from a haemorrhagic injury. The typical cardiovascular adjustments that occur during heat stress are an increase in heart rate resulting in an augmented cardiac output and peripheral vasodilatation, which reduces systemic vascular resistance. The net result is a displacement of blood from the central to the peripheral (skin) circulation to accommodate the skin blood flow requirements for effective heat transfer. While these adjustments generally do not have a negative impact on arterial pressure in the supine posture, the superimposition of central hypovolaemia as incurred during haemorrhage can quickly lead to a precipitous fall in arterial pressure, potentially leading to loss of consciousness and/or cardiovascular collapse. As such, understanding the cardiovascular physiology of haemorrhage in the heat stressed human has obvious clinical implications for the proper management of such scenarios. Recent advances point towards a shift in the operating point of the Frank–Starling mechanism (Wilson et al. 2009), reduced central blood volume (Crandall et al. 2008) and an altered capacity of the cutaneous vasculature to vasoconstrict (Wilson et al. 2002; Crandall et al. 2010), as a multifactorial mechanism for the compromised regulation of arterial pressure during (simulated) haemorrhage in the heat stressed human. In this issue of The Journal of Physiology, Crandall et al. (2012) reveal an intriguing component of the cardiovascular physiology of haemorrhage under heat stress. By combining scintigraphy (gamma camera imaging) with technetium-99m labelled red blood cells, Crandall et al. (2012) examined changes in central blood volume in heat stressed individuals during a simulated haemorrhagic challenge induced by lower body negative pressure. As passive heat stress itself reduces central blood volume (Crandall et al. 2008), the novel aspect of their study consisted of performing the simulated haemorrhagic challenge with (colloid volume infusion) and without prior restoration of central blood volume. Building upon previous research demonstrating that volume loading returned orthostatic tolerance during heat stress to levels similar to when normothermic (Keller et al. 2009), it was hypothesized that colloid volume loading sufficient to restore central blood volume to pre-heat stress levels would attenuate the reductions in central blood volume during a subsequent haemorrhagic challenge. Contrary to this hypothesis, however, decreases in central blood volume were similar whether volume loading was performed or not prior to the haemmorrhagic challenge. To reconcile the seemingly contradictory observations that volume loading can improve orthostatic tolerance, yet does not attenuate decreases in central blood volume, the authors nicely point out that warm blood vessels have a remarkable capacity to increase blood volume under an elevated state of hyperthermia. As such, the added central blood volume provided by colloid infusion is simply absorbed by the dilated vasculature of the lower limbs. Aggravating the problem is also a lower vasoconstrictor responsiveness of the heat stressed cutaneous vasculature (Wilson et al. 2002), which effectively keeps the infused volume within the ‘sink’ created by the heated lower limbs. That being said, volume infusion does elevate cardiac output to greater levels prior to initiating the simulated haemorrhagic challenge; such that volume loaded individuals maintain a greater cardiac output for a given level of lower body negative pressure while heat stressed. This is likely to be the mechanism by which volume loading can maintain orthostatic tolerance, despite having little effect on decreases in central blood volume. These findings raise the important question as to whether volume loading alone is an adequate intervention treatment to maintain both central blood volume and consciousness during haemorrhage under stress. It might be that greater infusion volumes than the one used in the current study are needed to significantly limit decreases in central blood volume in such scenarios. However, the capacity of warm blood vessels to increase blood volume might require the use of unreasonably large infusion volumes. In contrast, future studies might consider interventions that target both the decreases in central blood volume and the ‘sink’ created by the warm vascular beds. For example, skin surface cooling results in similar cardiovascular adjustments as colloid volume infusion, such as a rightward shift of the operating point of the Frank–Starling relationship toward a flatter portion of the curve (Wilson et al. 2009), resulting in greater orthostatic tolerance to a simulated haemorrhagic challenge. Skin surface cooling would also be expected to induce vasoconstriction of the peripheral vascular beds and re-distribute blood towards the central circulation. The distinct advantage of skin surface cooling, however, might also be to limit the accumulation of blood in the compliant venous system, possibly making a given volume loading more effective by keeping a greater proportion of it within the central circulation. Although body warming of haemorrhagic patients has long been advocated to avoid the complications of potential hypothermia, it must be kept in mind that the thermal and cardiovascular consequences of hyperthermia might allow for combined skin surface cooling and volume loading to be an effective treatment strategy in heat stressed individuals.