Firing costs, together with the legislative and regulatory frameworks governing employment relations, are often blamed for poor labour market outcomes. Yet, research on the economic impacts of these costs is inconclusive. There has been much focus on functional assumptions and the significance of parameters, but very little on the quality and precision of the cost measures upon which most results hinge. Reviewing the indirect and direct measurement methods commonly used, the author argues that direct quantitative methods, rarely used in research, are much needed to complement more popular indirect measures. A recent survey experiment conducted in Australia illustrates this point.