Variant of the case control study in which the controls are drawn from the same cohort as the cases but are identified before the cases develop; some of the controls may later become cases.