In a prospective cohort study, groups of people are identified before they show any signs of disease and are followed up over time. Alternatively, in retrospective cohort studies, data is used that has already been collected (possibly over a long period of time) for other purposes.
Cohort studies are useful when it would be unethical to carry out a randomised controlled trial (RCT). For example, it would be unethical to deliberately expose people to cigarette smoke or asbestos.