An experimental design structure which matches subjects as closely as possible and then assigns one of each pair to each of the treatment group and control group. This might involve cutting pieces of cloth in two when testing detergents, or pairing patients in a clinical trial by age, gender, weight, etc. as closely as possible. The allocation of treatments to the paired subjects should be randomized. Then the differences in outcome are measured. The strength of this design lies in reducing the amount of variation between subjects so that any actual differences due to the experimental conditions are more easily identified.