Despite increased stroke volume (SV), 'athlete's heart' has been proposed to have a similar left ventricular (LV) muscle function - as represented by LV twist - compared with the untrained state. However, the underpinning mechanisms and the associations between SV/cardiac output and LV twist during exercise are unknown. We hypothesised that endurance athletes would have a significantly lower twist-to-shortening ratio (TwSR, a parameter that relates twist to the shortening of heart muscle layers) at rest, but significantly greater LV muscle function during exercise. Eleven endurance trained male runners and 13 untrained males were tested at rest and during supine cycling exercise in normoxia and hypoxia (increased cardiac output but unaltered SV). Despite the expected cardiac remodelling in endurance athletes, LV twist, torsion, TwSR, strain and strain rate ('LV systolic mechanics') did not differ significantly between groups (P > 0.05). Structural remodelling, as per relative wall thickness, and LV twist did not correlate (r =Â 0.04, PÂ =Â 0.33). In normoxia and hypoxia, exercise increased LV systolic mechanics in both groups (P < 0.001), but with different relationships to SV and cardiac output. Conversely to our hypothesis, hearts of different size had similar LV systolic mechanics, suggesting that similar twist, torsion and TwSR at rest and during exercise irrespective of cardiac output may be an important mechanism in healthy hearts. We hypothesise that the regulatory 'purpose' of LV twist may be related to the sensing of maximal cardiac myofibre stress, which may act as a biologically purposeful limiter to contraction.