Missing data of concrete dam deformation can cause difficulty and even misjudgment to the analysis of dam service status. Based on the theory of clustering and panel data, a novel method to deal with the missing data of concrete dam deformation is proposed, in which the common and specific influencing factors of dam deformation are considered. Compared with traditional methods, this method can more comprehensively reflect the influence of load and non-load factors on dam structural performance, and is more effectively for processing the missing data of dam deformation. The application of missing data processing in a concrete double curvature arch dam shows that the predicted residual error is lower than the error limit of arch dam radial deformation stipulated in the Code for Dam Safety Monitoring (SL.601-2013), indicating that it has high precision and can effectively process the missing data of safety monitoring.