A bridge is an infrastructure that experiences dynamic and repetitive loading for a long period, so it is necessary to monitor the condition of the bridge structure to ensure that the bridge is always in proper condition. Deflection is the vertical displacement that can occur in the bridge. This study aims to obtain an correlation model in the form of regression equation with multiple regression analysis methods between the deflection value and the rotation value of the PCI Girder bridge. The analysis was carried out using Finite Element Method software, by modeling the PCI Girder bridge with several vehicle load cases, and the rotation and deflection values of the bridge will be obtained. Several schemes were made regarding the number and location of rotation sensors to determine the level of accuracy. To get the relationship between rotation value and deflection value, a multiple linear regression analysis was carried out. It can be generated a multiple linear regression equation with the dependent variable being the deflection value and the independent variable being the rotation value on the bridge placement. This equation can be used to predict the deflection value in the middle of the bridge span with the percentage difference shows value smaller than 5%. In the analysis, it is able to provide results with the smallest percentage difference between numerical comparison values and linear regression equations.