This article shows the relevance of Burawoy’s analytical categories on the division of scientific labour to analyse the interactions between science and practice. It argues that social scientists need to combine the roles of expert witness, social engineer and storyteller in order to develop a productive relationship with policy makers. It also emphasises the relevance of a permanent dialogue between social scientists and policy makers. However, the analysis disagrees with Burawoy’s view that knowledge is based on a consensus between scientists and their publics (consensual knowledge). Burawoy underestimates the risk of a politicisation of science inherent to a close relationship with various ‘publics’. The arguments presented in this article are based on the reception of a research project on labour migration from Central and Eastern Europe. This was a hybrid research project in which different actors participated: the Erasmus University, nine Dutch municipalities, the national government and a Dutch knowledge center on urban issues.