Objective: To study an epidemiological methodology for domestic violence. Methods:A program was drawn,the epidemiological instruments were developed, investigators were trained, not only according to the principles of epidemiology but also being suited to Chinese condition. Then,We win over the support of government bodies, arouse the messes to action, and all clues from investigation were confirmed by interview with persons concerned.Spot check randomly was used for control of quality.Results:The reliability of the epidemiological instruments was 97%, the sensitivity was 95%, the specificity was 100%. The prevalence of domestic violence in life-time and past year was 16.2% and 11.6% respectively.Conclusion:This was a practical methodology in epidemiological survey for domestic violence in China.