We examined the hypothesis that changes in the suicide rate of England and Wales since 1900 are an artefact of changes in the process of classifying violent deaths. Four potential sources of artefact were considered: (1) the reliability of the procedure for investigating suspected suicides; (2) coding by the Office of Population Censuses and Surveys; (3) changes in the law; (4) High Court decisions on suicide. None of these can explain the large observed fluctuations in the suicide rate.