Considering that displacement monitoring index of concrete dam is affected by the sensitivity of structural monitoring time series and performance degradation，in order to improve the reliability of the displacement monitoring index on the early-warning function for safety behavior of concrete dam service，a method for determining displacement monitoring index of concrete dam based on the hybrid model considering parameter interval inversion modification is carried out on the combination of interval analysis theory，rough theory and neural network theory in the paper. The hybrid model with the function of interval parameter inversion is established on the basis of dam prototype monitoring data，which combines rough theory and neural network theory. The material parameter intervals of concrete dam and bedrock are inversed by the model. Concrete dam displacement monitoring index is determined by the most unfavorable combination of parameters inversed by interval inversion. Examples show that the method combines the uncertain data excavation ability of rough theory and self-learning ability on nonlinear problem of neural network，it has good robustness. Material parameter intervals of concrete dam inversed by the method are reasonable and effective，the monitoring index carried out by the method provides a theoretical basis and decision support for long service and operation management of concrete dam. In addition，the method can be used for determining displacement monitoring indexes of other structures after some expansion and improvement.