In recent years, Post-Keynesian analysis has been characterized by a renewed interest in long-run theories of growth and distribution. While many authors have focused on the convergence of demand-led growth models to a fully adjusted equilibrium, relatively little attention has been given to the time required to reach this long-run position. In order to fill the gap, this paper seeks to answer the question of when is the long run in demand-led growth models. By making use of numerical integration, it analyses the time of adjustment from one steady-state to the other in two wellknown demand-led growth models: the Sraffian Supermultiplier and the fully adjusted version of the neo-Kaleckian model. The results show that the adjustment period is generally beyond an economically meaningful time span, suggesting that researchers and policy makers ought to pay more attention to the models' predictions during the traverse rather than focusing on steady-state positions.