Scientific technique designed to eliminate bias in a study.  It does this by randomly placing the subjects into experimental (receives treatment) and control groups (receives no treatment) and neither the subject nor the researchers known which group is receiving treatment(s) and which is not.