Odds are another way to measure the association between exposure and outcome. It is particularly useful in case-control studies. The most common measure related to odds is the odds ratio (OR), which compares the odds of exposure in cases to the odds of exposure in controls.