This article presents evidence on the link between employment protection legislation (EPL) and the rate of unemployment in a cross‐country panel dataset of Organisation for Economic Co‐operation and Development countries from 1985 to 2013. We use both a traditional panel specification with lags of the policy variable, and also a unique structural panel vector autoregression (PVAR) method to determine the long‐run dynamic interaction between EPL and unemployment. We confirm that more restrictive EPL for permanently employed workers causes a significant and persistent increase in unemployment, but the effect is only apparent at long‐lag lengths, some 2–5 years after the law has been implemented. (JEL J68, J65, J63)