The contemporary international relations literature links the democratic peace hypothesis to Kant’s famous peace plan. Yet, whether attempting to prove or disprove the hypothesis, most quantitative studies have lost sight of important dimensions of the Kantian vision. I reinterpret the democratic peace as a dynamic and dialectical learning process. In order to assess the dynamic dimension of this process (while controlling for exogenous dialectical reversals), I rely on quantitative evidence drawn from popular data sets. In conformance with the Kantian perspective, the conflict propensities among democracies exhibit a steadily falling trend since the nineteenth century. Yet, in partial opposition to Kant’s expectations, other dyads also experience a significant, although weaker, pacifying trend. A series of tests shows that these findings are robust to epochal effects, various control variables, and “maturity effects” measuring the age of democratic dyads.