Used in randomised controlled trials. A computer decides at random which treatment or treatment strategy a trial participant will receive. This ensures that each participant has the same chance of receiving the treatments or strategies being compared, and avoids one treatment being given to someone because they are, for example, elderly or very sick. Randomisation ensures that the groups of people being compared in a trial are as similar as possible, except for the treatment they receive. This in turn ensures that differences seen between these groups after they have started their treatment are likely to be due to the treatments being compared.