In their proposal on a “Legislature by lotâ€, John Gastil and Erik Olin Wright (2017) propose to launch a bicameral parliament where one chamber would be composed by elected politicians and the other by sortitioned ordinary citizens. Their idea is to combine party-independent problem-solving deliberations in a chamber for which every citizen has the same chance to get selected, with adversial debates on conflicting policy preferences in a chamber for which every citizen was entitled to cast a vote. Given that election and sortition have distinct virtues, this looks like a promising proposal to reintroduce sortition in the selection of political representatives. Now, if the authors envision the interaction between the two chambers as a “creative tensionâ€, it seems to us that this question has received less attention than it would deserve. The assemblies have not only different virtues but also different legitimacies which can become particularly conflictual if each chamber has the power to veto the proposals of the other, as Gastil and Wright recommend. If the elected chamber is less popular than the sortitioned one, the legitimacy of the election mechanism might come to be more and more questioned. In turn, elected representatives might try to bring discredit upon the allotted representatives. The conflict would thereby not only occur during the legislative process, but also in public opinion. We propose to address this possible tension between the elected and the sortitioned chamber both theoretically and empirically. In Belgium, several propositions have been made by politicians, academics and intellectuals to transform the Senate (the upper house) in a sortitioned assembly. Thereby, we use quantitative and qualitative data to analyse the opinion of both politicians and citizens on a sortitioned chamber. After exposing the complementary virtues and competing legitimacies of election and sortition from the point of view of political theory, we draw back on the reactions to the propositions in Belgium by ...