Congestion is prognostically relevant in cardiac transthyretin amyloidosis (ATTR-CA), but whether congestion has an incremental prognostic value beyond the well-established, congestion-sensitive NT-proBNP is unknown. Therefore, we aimed to comparatively evaluate the prognostic utility of several congestion surrogates over NT-proBNP. We estimated hazard ratios by Cox proportional hazards regressions with time-varying covariates from a panel data set of the local amyloidosiscohort study AmyKoS. Different models were compared by using chi(χ)2-statistics measuring overall model significance. 131 ATTR-CA patients (wild-type 84.0%, hereditary 6.9%, without genetic testing 9.2%; median age 78.7 (quartiles 73.3, 82.1) years; 85.5% male) with 566 observations across a median follow-up of 38.2 (30.6; 48.2) months were analyzed. 83.2% received disease-modifying treatment; 20.6% participated concurrently in placebo-controlled gene silencer trials. Information on congestion improved biomarker-driven risk stratification and identified patients at the highest risk. Echocardiographic congestion markers performed better than clinical findings and daily diuretic use/dosage. Relevant adjusters were daily diuretic dosage, disease-modifying treatment, eGFR, and right atrial volume. NT-proBNP and the tricuspid regurgitation peak velocity (tr-vmax) provided an easy-to-use stratification with overall model performance similar to NAC and Mayo staging systems. Further analyses are necessary for validation and to identify the optimal cut points of the congestion markers.