This paper presents the results of a numerical study on the predictive capabilities of Large Eddy Simulation (LES) and hybrid RANS/LES methods for heat transfer in the trailing edge (TE) geometry experimentally investigated in Part 1. The experimental validation data includes 2D wall contours and laterally-averaged values of adiabatic cooling effectiveness. The simulations were conducted at three different blowing ratio values. The comparison with the experimental data shows a general advantage of LES and hybrid RANS/LES methods against steady-state RANS. The results obtained by means of the WALE LES model and the Improved Delayed Detached Eddy Simulation (IDDES) hybrid RANS/LES method were comparable. The presented grid dependence study shows the importance of adequate grid resolution for the predictive capabilities of trailing edge cooling LES. Furthermore, the importance of considering TE slot lands simulation quality in the numerical method assessment is discussed. Potential directions of future research needed to improve simulation reliability are outlined.Copyright © 2015 by ASME