Abstract. In turbomachines, transitional ﬂows are likely to occur over many components. In the case of a low-pressure turbine, the transition from laminar to turbulent boundary layer is often separation-induced. The overall blade losses depend among others physical eﬀects on the size and the length of the separation bubble. The characteristics of the separation bubble are strongly linked with the Reynolds number and the turbulence intensity of the ﬂow. So, turbomachinery designers require tools which could be used to achieve an accurate prediction of the laminar-turbulent transition. Many computational techniques are available to predict transitional behavior. However, only the computation of the Reynolds Averaged Navier-Stokes (RANS) equations seems to be able to be used in an industrial context, due to the relatively low computational ressource required. The aim of this paper is to present and compare diﬀerent approaches used to predict transition on practical test cases. Two models using experimental correlations to simulate transitional ﬂows are described. The ﬁrst uses boundary layer integral values, and the second uses only local values. The chosen test cases are two low-pressure turbine cascades with diﬀerent aerodynamic loads and a multistage low-pressure turbine.