Exercise can lead to release of biomarkers such as cardiac troponin T (cTnT) and N-terminal pro-brain natriuretic peptide (NT-proBNP), a poorly understood phenomenon proposed to especially occur with high-intensity exercise in less trained subjects. We hypothesised that haemodynamic perturbations during exercise are larger in athletes with cTnT release, and studied athletes with detectable cTnT levels after an endurance event (HIGH; n = 16; 46 ± 9 years) against matched controls whose levels were undetectable (LOW; n = 11; 44 ± 7 years). Echocardiography was performed at rest and at peak supine bicycle exercise stress. Left ventricular (LV) end-systolic elastance (E (LV) a load-independent measure of LV contractility), effective arterial elastance (E (A) a lumped index of arterial load) and end-systolic meridional wall stress were calculated from cardiac dimensions and brachial blood pressure. Efficiency of cardiac work was judged from the ventriculo-arterial coupling ratio (E (A)/E (LV): optimal range 0.5-1.0). While subgroups had similar values at rest, we found ventriculo-arterial mismatch during exercise in HIGH subjects [0.47 (0.39-0.58) vs. LOW: 0.73 (0.62-0.83); p < 0.01] due to unopposed increase in E (LV) (p < 0.05). In LOW subjects, a greater increase occurred in E (A) during exercise (+81 ± 67 % vs. HIGH: +39 ± 32 %; p = 0.02) which contributed to a maintained coupling ratio. Subjects with higher baseline NT-proBNP had greater systolic wall stress during exercise (R (2) = 0.39; p < 0.01) despite no correlation at rest (p = ns). In conclusion, athletes with exercise-induced biomarker release exhibit ventriculo-arterial mismatch during exercise, suggesting non-optimal cardiac work may contribute to this phenomenon.