Often the accuracy or reliability of the time stamps of power quality data coming from different measurement locations is not adequate for certain analysis tasks. This paper studies improving the accuracy and reliability of the time stamps by utilizing the mutual correlations of the measured signals. Especially the electricity grid fundamental frequency estimated using a reliable and stable high performance adaptive method is almost the same everywhere in a synchronized power system. That enables accurate identification of clock time differences and drifts. The paper compares the results with the time stamps given by time managed local metered data preprocessing and concentrator servers. The results show that the identified time differences agree well with the time stamps given by the servers, but also reveals that the identified timing difference makes it possible to detect and correct the large timing errors that occasionally appeared in the time stamps added by the server.