Auditors in Medicare overpayment or False Claims Act (FCA) cases often use statistical extrapolation to estimate a health-care provider’s total liability from a small sample of audited claims. Courts treat statistical extrapolation differently depending on the context. They generally afford the government substantial discretion in using statistical extrapolation in overpayment cases. By contrast, courts typically more closely scrutinize the use of extrapolation in FCA cases involving multiple damages and civil penalties to ensure that the sample truly reflects the entire universe of claims and that the extrapolation rests on a sound methodological foundation. In recent cases, however, multiple courts have allowed the use of extrapolation in FCA cases. When auditors attempt to use statistical extrapolation, providers should closely inspect the sample and challenge the extrapolation when any reasonable argument exists that the sample does not constitute a reliable or accurate representation of all the provider’s claims.