Case control or retrospective studies begin by identifying individuals who do and do not have the disease under investigation. It is important to consider how these individuals are identified to avoid potential selection bias. Then we determine which individuals were exposed to a particular risk factor by asking them to recall the exposure, going back into medical records where the exposure may have been documented, or using a biomarker to measure the exposure. These different methods of documenting exposures may not be equally accurate.

Example:
An investigator identifies 100 individuals with multiple sclerosis (MS) (cases) and 100 individuals who do not have MS (controls). Controls are often group-matched by age, gender and race so they are comparable to the cases. Ten of the cases had been exposed to a particular toxin during childhood. Only 1 of the controls reported a similar exposure. The prevalence of the exposure among cases and controls is 10% and 1%, respectively. Because this is statistically significant, this exposure is considered to be a risk factor for MS.