Resting right ventricular (RV) systolic function has in some studies been shown to be impaired after correction of an atrial septal defect (ASD) whereas impairment of left ventricular (LV) systolic function is uncertain. In the present study we examine the LV and RV systolic response to exercise in patients with a previously corrected ASD in order to investigate the myocardial capacity. Thirty-six adult ASD patients with a corrected isolated secundum ASD and eighteen adult age-matched controls underent a semi-supine exercise stress echocardiographic examination. At rest, LV parameters were comparable between groups, and RV global longitudinal strain (RV-GLS) was lower for the ASD group (-18.5%, 95% CI -20.0--17.0%) compared with controls (-24.5%, 95% CI -27.7--22.4%, p < 0.001). At peak exercise, LV ejection fraction (LVEF) was lower for ASD patients (61%, 95% CI 58-65%) compared with controls (68%, 95% CI 64-73% p = 0.01). Peak LV global longitudinal strain (LV-GLS) was borderline significantly lower (ASD: -18.4%, 95% CI -20.2--16.6%, controls: -21.3%, 95% CI -23.6--19.0%, p = 0.059). Both RVEF (ASD: 64%, 95% CI 60-68%, controls: 73%, 95% CI 65-80%, p = 0.05) and tricuspid annular plane systolic excursion (TAPSE) (ASD: 2.5 cm, 95% CI 2.3-2.7 cm, controls: 3.2 cm, 95% CI 2.9-3.6 cm, p < 0.001) at peak exercise were lower for ASD patients. Exercise assessed peak oxygen uptake was comparable between groups (ASD: 32.8 mL O2/kg/min, 95% CI 30.3-35.5 mL O2/kg/min, controls: 35.2 mL O2/kg/min, 95% CI 31.6-38.8 mL O2/kg/min, p = 0.3). Corrected ASD patients demonstrate a reduced LV and RV systolic exercise response decades after ASD correction whereas resting parameters of LV and RV systolic function were within normal range. The presence of subclinical systolic myocardial dysfunction during exercise might be associated with the long-term morbidities documented in this patient group.