A research design in which cases of a given disease or other condition are sampled for study. A comparison sample of persons not having the condition is also recruited, as far as possible from the same basic population. These samples are then compared on exposure to hypothesised risks for the condition.