Abstract: A perfect clinical trial must have a solid study design, strict conduction, complete quality control, non-interference of statistical result, and acceptable risk-benefit ratio. To reach the target, the quality control (QC) should be performed from the study design to conduction, from the analysis to conclusion. We discuss the relationship between data management and biostatistics from the statistical point of view, and emphasize the importance of the statistical concept and methods in the improvement of data quality in clinical data management.