A physiological control algorithm is being developed to ensure an optimal physiological interaction between the ReinHeart total artificial heart (TAH) and the circulatory system. A key factor for that is the long‐term, accurate determination of the hemodynamic state of the cardiovascular system. This study presents a method to determine estimation models for predicting hemodynamic parameters (pump chamber filling and afterload) from both left and right cardiovascular circulations. The estimation models are based on linear regression models that correlate filling and afterload values with pump intrinsic parameters derived from measured values of motor current and piston position. Predictions for filling lie in average within 5% from actual values, predictions for systemic afterload (AoPmean, AoPsys) and mean pulmonary afterload (PAPmean) lie in average within 9% from actual values. Predictions for systolic pulmonary afterload (PAPsys) present an average deviation of 14%. The estimation models show satisfactory prediction and confidence intervals and are thus suitable to estimate hemodynamic parameters. This method and derived estimation models are a valuable alternative to implanted sensors and are an essential step for the development of a physiological control algorithm for a fully implantable TAH.