Abstract Funding Acknowledgements Type of funding sources: None. OnBehalf MASCOT investigators Background Few studies analyzed left atrial (LA) peak atrial longitudinal strain (PALS) determinants, particularly across heart failure (HF) stages. We aimed to analyze the pathophysiological and clinical PALS correlates in a large international prospective registry. Methods This is a multicenter prospective observational study enrolling 745 patients with HF stages 0-C from July to October 2018. Data included PALS and left ventricular global longitudinal strain (LV-GLS). Exclusion criteria were: valvular prosthesis; atrial fibrillation; cardiac transplantation; poor acoustic window. Results Median global PALS was 17% [24-32]. 29% of patients were in HF-stage 0/A, 35% in stage-B, and 36% in stage-C. Together with age, the echocardiographic determinants of PALS were LA volume and LV-GLS (overall model R²=0.50, p < 0.0001). LV-GLS had the strongest association with PALS at multivariable analysis (beta:-3.60 ± 0.20, p < 0.0001). Among HF-stages (Figure 1), LV-GLS remained the most important PALS predictor (p < 0.0001) whereas age was only associated with PALS in lower HF-stage 0/A or B (R=-0.26 p < 0.0001, R=-0.23 p = 0.0001). LA volume increased its association to PALS moving from stage 0-A (R=-0.11; P = 0.1) to C (R=-0.42; P < 0.0001). PALS was the single most potent echocardiographic parameter in predicting HF stage (AUC for B vs. 0/A 0.81, and AUC vs. 0/A for C 0.76). PALS remained independently associated with HF stages after adjusting for ejection fraction, E/e’ ratio and mitral regurgitation grade (p < 0.0001). Conclusion Although influenced by LV-GLS and LA size across HF-stages, PALS is incrementally and independently associated with clinical status. LA function may reflect a substantial part of the hemodynamic consequences of ventricular dysfunction. Abstract Figure 1