AbstractUsing residual displacement as a measurable indicator can reduce subjectivity and uncertainties in postearthquake assessments. However, the accuracy of analytical models in predicting displacements, particularly residual displacement, has always been an obstacle in this field. This paper aims to fill the gap in the literature on a proper method to enhance the analytical predictions. A comprehensive parametric investigation on the accuracy of distributed plasticity models was carried out, with a focus on bridges under near-field motions. These parameters include influential nonlinear modeling parameters, loading history, viscous damping, and strain rate. The results of the 20 considered models are compared with the six available shake table test results, and two methods are proposed to enhance the displacement predictions for practical cases. In the first method, actual maximum and residual displacements can be determined with the desired confidence level. In the second method, nonlinear regression analysis based on many influential parameters is conducted to enhance the accuracy of analytical residual displacement. This method considerably reduces the model error compared to the available methods in the literature.