Abstract Funding Acknowledgements Type of funding sources: None. Background Global longitudinal shortening (GLSh) 1,2 and left ventricular ejection fraction (LVEF) are different measures of left ventricular function. LVEF is well-established, but can miss early or mild disease. GLSh, which quantifies long-axis shortening, is easy to measure, and has been shown to have superior reproducibility and prognostication ability than global longitudinal strain3,4. Purpose To assess the relationship between GLSh and EF measured using CMR in a large cohort representing health (healthy volunteers, athletes) and disease (hypertension, aortic stenosis, hypertrophic cardiomyopathy(HCM),cardiac amyloidosis, Fabry disease). Methods CMR scans from a study cohort of 932 subjects were analysed (121 healthy volunteers, 278 athletes, 40 patients with hypertension, 76 severe aortic stenosis, 164 HCM, 51 cardiac amyloidosis, 202 Fabry). EF and GLSh were automatically measured from standard CMR SSFP cine images using machine learning methods described previously3,5. Correlation coefficients between GLSh and LVEF were calculated for each disease category. Normal reference ranges for LVEF and GLSh were estimated from a subset of n=4,342 healthy subjects from the UK Biobank, who had no history of cardiovascular disease or medication use. Results The lower limit of normal in the UK Biobank healthy subjects was 55% for EF and 13% for GLSh. Across the study cohort, for subjects with normal LVEF, only a small minority of healthy volunteers (n=1, 0.8%) and athletes(n=2; 0.7%) had low GLSh, whereas rates increased progressively in hypertension(n=3;9%), Fabry(n=61;31%), aortic stenosis(n=26;44%), HCM (n=87; 56%), and cardiac amyloidosis(N=28; 93%). Conversely, few subjects (n=8/932;0.1%) have the opposite: normal LGSh but low LVEF(Figure 1). The relationship between LVEF and GLSh is shown in Figure 2. Apart from HCM (r=0.019,p = 0.8), all other diseases show a significant association between GLSh and LVEF, with progressively higher correlation from healthy volunteers (r=0.33 p<0.001) to cardiac amyloidosis (r=0.33; p = 0.0016), athletes (r=0.44; p<0.001), and hypertension (R=0.6; p<0.001), with aortic stenosis showing highest correlation (r=0.68,p<0.001). Fabrys patients showed the reverse trend, with an inverse correlation between GLSh and LVEF (r=-0.34; p<0.001). Conclusions In health, impaired GLSh is rare when LVEF is normal, however across diseases with hypertrophy, GLSh impairment is common, illustrating the potential of GLSh to detecting early or mild systolic impairment. There are unexpectedly different associations between GLS and EF. In some diseases, these appears to measure the same aspects (health/AS/hypertension), but in others they are not well correlated with GLS providing new information. In Fabry, there is an unexpected inverse relationship. Further work is needed to understand this.