Data management in clinical trials is an activity performed with high ambitions but under undefined regulatory requirements as to the quality of the data presented in the trial reports. The term quality is not defined but its use implies a degree of absence from measurement errors. The term error is not defined but its use indicates an unexplained difference between the original recording of measurement and data presented in the final report. From a review of modern quality management methods in comparison to standard procedures in clinical data management it is apparent that there are striking differences in many aspects on how to assure high quality. On basis of this review it is suggested that the major obstacle for obtaining high quality is the traditional dependence on repeated reviews of inconsistent data late after the original measurements have been made. More systematic use of experience and resources, in particular for training investigators, should be made to prevent errors.