In vitro cardiac myofibre contraction and relaxation are linearly associated across a wide range of muscle fibre (or sarcomere) lengths, contraction frequencies and β-adrenergic activation (Janssen, 2010, see Fig. 1A). Whether linearity of cardiac contraction-relaxation coupling is maintained when in vivo O2 availability is reduced and the heart muscle itself experiences a greater demand for coronary flow remains to be elucidated. Therefore, in a within-day repeated-measures experiment, sixteen healthy males (age: 22 ± 4 years, VO2peak: 45.5 ± 6.9 ml/min/kg) exercised at i) 2% above individual anaerobic threshold (IAT) and ii) 40% of normoxic peak power in room air (40%peak/NORM) and iii) in hypoxia (40%peak/HYP, FiO2 = 12%). Left ventricular (LV) contraction and relaxation were determined from speckle-tracking ultrasound-derived twist and untwisting rate, respectively. From rest to exercise at 40%peak/NORM and IAT, LV twist and untwisting rate increased linearly (y = -8.7x) in accordance with the aforementioned literature (Janssen, 2010, see Fig. 1A). However, during exercise at 40%peak/HYP, LV untwisting rate was disproportionally increased compared with LV twist (y = -18.1x, see Fig. 2B). This phenomenon occurred while whole-body O2 consumption was similar between 40%peak/NORM and 40%peak/HYP (mean ± SD, 22.3 ± 4.9 and 22.2 ± 4.6 mL∙kg-1∙min-1, respectively, P > 0.05) but rate pressure product, a surrogate of cardiac O2 consumption, was significantly increased in hypoxia (21124 ± 2313 and 26823 ± 3309 bpm∙mmHg-1, respectively, P < 0.001). The disproportionate increase in LV untwisting rate was not associated with the preceding twist rate, frequency of contractions or altered LV volumes. In conclusion, the present data provide in vivo evidence of an acute uncoupling of LV contraction and relaxation as represented by twist and untwisting rate. We hypothesise that LV contraction and relaxation may be regulated at least in part separately, according to the peripheral whole-body vs. coronary O2 demands. This finding may help to explain pathological conditions in which systolic and diastolic function are affected differentially, for example in “diastolic heart failure” (Zile et al., 2004), also termed “heart failure with preserved ejection fraction” (Andersen & Borlaug, 2014).