A trial design in which neither the researcher, the person administering the therapy nor the patient knows the details of the treatment the patient was randomly assigned to.