Scholars working on democratic peace theory (DPT) often think of their work as nothing but a rejuvenation of Kant's insights, backed by empirical data only available to contemporary observers with a vantage view point of two centuries of historical events. Kant's theory of peace and DPT, however, are two very different models and marking the difference is the central aim of this paper. We aim to show that each of the three pillars on which Kant's theory rests (the three definitive articles) has been misinterpreted by DPT scholars. We also intend to show that Kant's model is superior to DPT from a normative standpoint, i.e. it offers better guidance for progressing towards a more peaceful world than the model based on the 'separate peace' dear to Michael Doyle and his followers. Given these two goals, the paper naturally falls into two main parts. The first part criticizes the way DPT scholars read Kant and marks the distance between the two models. The second part shows how the Kantian model, now clearly distinguished from DPT, articulates a path for the pacification of international relations that is considerably more attractive than the model suggested by DPT.