This paper analyses the evolution of the European urban system from a long‐term perspective (from 1300 to 1800). Using the method recently proposed by Clauset, Shalizi, and Newman, a Pareto‐type city size distribution (power law) is rejected from 1300 to 1600. A power law is a plausible model for the city size distribution only in 1700 and 1800, although the log‐normal distribution is another plausible alternative model that we cannot reject. Moreover, the random growth of cities is rejected using parametric and non‐parametric methods. The results reveal a clear pattern of convergent growth in all the periods.