Why do democratising nations make the constitutional choices they do? Conceiving democratic transition as a critical juncture, we propose a theory of constitutional choice. We place the degree of uncertainty at the centre of our theorising efforts to explain the relationship between constitutional bargains among competing political groups and the type of executive–legislative relations adopted during democratisation. We posit that parliamentarism is more likely to be adopted under high-uncertainty conditions, while presidentialism is more likely under low-uncertainty conditions. Identifying four factors that affect the level of uncertainty in the transition process, we examine how the choices of executive–legislative relations are made under strong influences of historical and geographic factors.