I contrast the effect of plurality voting and proportional representation (PR) on candidates’ entry behavior. I do so using the citizen‐candidate paradigm (Osborne and Slivinski, 1996 and Besley and Coate, 1997). To this aim the paper introduces a new way of modeling PR that takes coalitions explicitly into account and compare the equilibria to those under plurality voting and PR without coalitions (Hamlin and Hjortlund, 2000). I find that (i) taking coalitions into accounts reduces candidate polarization; (ii) for policy‐motivated candidates PR leads to more polarized entrants; (iii) for office‐motivated candidates PR with coalitions is most conducive to multi‐candidate equilibria. I complement the theoretical analysis with data from a laboratory experiment comparing plurality voting and PR without coalitions. In line with theoretical predictions I find that higher costs of entry lead to fewer entrants that are more centrist and that PR leads to weakly more entry than plurality. Field: Experimental Economics