This review compared seven distinct methods of identifying medication errors in hospitalized patients—including chart review, incident reports, direct observation, and review of administrative data—but concluded that insufficient evidence exists to support routine use of one method over another.