Purpose – The purpose of this paper is to examine whether works councils (WCs) in Belgium have a positive effect on firm performance, notably productivity and profitability, while taking the role of trade unions into account. Design/methodology/approach – The authors first introduce the typical Belgian industrial relations system, discussing the similarities and differences with neighboring countries. This is followed by a brief overview of the relevant literature. Subsequently, the impact of Belgian employee representation on firm performance is estimated by means of OLS, using a newly developed questionnaire administered among Belgian CEOs. Special attention is given to moderating and mediating effects. Findings – The authors find that Belgian WCs have a small (direct) significantly positive effect on labor productivity, but not on profitability. The additional results of the mediation test show tentatively that WCs might affect profitability indirectly, through their impact on productivity. Despite trade unions’ dominance in practice, the findings reveal that their impact is insignificant. Research limitations/implications – Although nationwide, rich and representative, as well as statistically valid, the data set is rather small (196 usable observations). The data set offers ample opportunities to further explore what makes effective Belgian WCs different from their non-effective counterparts. Originality/value – The data set is unique, and combines subjective CEO with objective performance data. The data offer the opportunity to do a first study into the special case of Belgium, which has a distinct Union dominated IR regime. In this study, the focus is furthermore on the rarely studied WC-trade union interaction. In addition, subtle moderation and mediation effects are estimated.