In medical research, ensuring that the group of actively treated subjects has as many relevant similarities as possible to a group of untreated or placebo-treated persons. Matching subjects with controls increases the likelihood that the findings demonstrated by the study are the result of the treatment itself and not another variable.