Abstract Background/Introduction Amongst various potential biomarkers associated with cognitive deficits, phosphorylated Tau (pTau) appears to link brain and cardiac abnormalities through a systemic tauopathy pathway. Purpose A recent report on the potential importance of myocardial pTau in chronic heart failure (HF) prompted the here described longitudinal quantification of circulating pTau serum levels in patients with HF. We hypothesized that serum pTau accumulates in HF and is associated with markers of myocardial dysfunction. Methods Eligible patients within the prospective, investigator-initiated, monocentric follow-up study Cognition.Matters-HF were older than 18 years, suffered from stable chronic HF and were free from major neurological or psychiatric disorders. Beyond the cardiological assessment including transthoracic echocardiography and six minute walking distance (6-MWD), investigations at baseline, 1 year and 3 years comprised extensive, domain-specific psychological testing and volumetric brain magnetic resonance imaging. Serum pTau was measured using the pTau181 advantage kit. Results Serum samples were available from 78 patients. Mean age was 63±10 years and 87% were men. Serum pTau longitudinally increased from a median of 1.05 (quartiles 0.26, 1.21) pg/ml at baseline to 1.84 (1.36, 3.60) pg/ml after 3 years (p<0.001). During that time period, LVEF, NT-proBNP, 6-MWD and the hippocampal volume remained unaltered, whereas global brain volume and cortical thickness shrank age-appropriately. Cognitive testing revealed that domain-specific T-scores within 36 months remained stable for attentional intensity and executive functioning, and even improved for memory. When we entered baseline values of age and above-mentioned cardiac, brain-volumetric, and cognitive parameters into a multivariable regression model predicting log-transformed pTau, only age (β=0.34), NT-proBNP (β=0.24) and GBV (β=0.26) were independently associated with pTau serum levels (R2=0.18). Consistently, at 36 months, the same predictors were selected by the model (R2=0.29): age (β=0.41), NT-proBNP (β=0.43), and global brain volume (β=0.26). Moreover, higher baseline pTau (β=0.23) and lower 6-MWD (β=-0.39) independently predicted ΔpTau with an explained variance (R2) of 0.24. Conclusions We observed a prominent increase of serum pTau over time, which was predicted by prevalent myocardial dysfunction. Moreover, absolute NT-proBNP levels were independently associated with absolute pTau serum levels, along with age and global brain volume. Mechanistically, phosphorylation of cardiac tau is thought to destabilize its interaction with microtubules causing microtubule detyronisation and increased myocardial stiffness through enhancing the mechanical resistance of contracting cardiomyocytes. The etiological role of pTau in the development of HF and the potential of serum pTau as a novel biomarker of chronic HF warrants further studies.