Computational modelling of cardiovascular flows has become an essential method for haemodynamic analysis and clinical decision support [1]. Nevertheless, the accuracy of model predictions is constrained by uncertainties in outflow boundary conditions, where Windkessel models are frequently used to represent distal vascular impedance [2]. Although widely adopted, Windkessel parameters are seldom known in advance and can substantially affect predicted flow patterns, pressure distributions, and calculated haemodynamic metrics. This presentation introduces an uncertainty-aware data assimilation (DA) framework for calibrating Windkessel outflow parameters in unsteady incompressible cardiovascular flow simulations. Blood flow is governed by the unsteady incompressible Navier–Stokes equations and solved using a stabilised finite element approach [3], with Windkessel models implemented at the outlets to represent downstream vascular effects. Calibration is performed within a probabilistic DA framework [4], in which outflow parameters are modelled as stochastic variables and constrained by noisy flow and pressure observations. Unlike deterministic inverse approaches [5], this framework provides confidence intervals on calibrated parameters and propagates uncertainty to haemodynamic predictions. Validation using synthetic observations generated on patient-specific geometries demonstrates accurate parameter recovery and well-calibrated uncertainty bounds. Sensitivity analyses further reveal how parameter uncertainty affects clinically relevant metrics, including wall shear stress, pressure gradients, and outlet flow distribution. Altogether, the approach enables uncertainty-aware haemodynamic prediction and improves the robustness and reliability of patient-specific cardiovascular flow simulations.