In epidemiology, assigning people to treatment and control groups in an unbiased manner so as to produce groups similar relative to their characteristics, random sample.
In research, a method used to assign subjects to experimental groups without introducing biases into a study.