A study that looks backwards in time, i.e. using data on events that have already occurred when the study starts.