This paper analyzes critically the early and recent literature on earnings management. I summarize the motivations and estimation models for accrual earnings management and real earnings management. Specifically, I focus on estimation models in real earnings management which are developed by Vorst (2016), extending the original sample period. Cross-sectional analysis reveals that earnings management is detectable purely based on companies’ financial data. The analyses are of interest to investors, regulators, and researchers with respect to the identification and consequences of earnings management.