Understanding what a reviewer wants helps authors write better papers and design better research studies. Central to any research study is the research question. This is the first thing the reviewer looks for. The question, or research aim, should be relevant, original and very clearly defined. The paper should also be scientifically valid; the methods should achieve the aim by addressing the question, and the conclusions should match the results. The statistical analysis often attracts criticism, especially the relatively simple problems arising with the use of P-values and the concepts of statistical and clinical significance. There are several agreed guidelines describing exactly how studies should be reported. Following these is a great help to reviewers.