AbstractThis study investigates the effects of the mandatory extension of collective agreements on union membership. This relationship is investigated using a difference‐in‐differences framework on the staggered industry‐ and county‐wide introduction of mandatory extension in Norway from 2005 to 2011. The introduction of mandatory extension was championed by labour and social partners and motivated by the EU enlargement of 2004 in response to increasing labour immigration. However, mandatory extension could lead to lower union membership through free‐rider behaviour. It is therefore a question of whether the cure is worse than the problem it is supposed to solve. The results show that mandatory extension had an overall negative effect on union density in the affected industries of about −2.7 percentage points. The effect varies between the different industries but is never positive. These findings give credence to the free‐rider hypothesis, where mandatory extension is expected to lead to lower union density because there is less incentive to become a union member when workers can obtain the benefits of the collective agreement without paying the union dues. In line with the theoretical expectations, the effect is stronger where the gap between collective wage and non‐union wage is largest. However, negative effects are also present where this wage gap is small, showing that the results are valid in different contexts.