In gas turbine engines, the endwall misalignment (step geometries with various step heights) commonly exists between the combustor exit and the first nozzle guide vane endwall, due to assembly error, metal corrosion, and thermal expansion. The step geometries affect the secondary flow field near the endwall and coolant flow behavior, thereby leading to a significant variation in endwall film cooling, vane surface phantom cooling, and vane passage aerodynamic performances. This paper presents a detailed experimental and numerical study on endwall film cooling and vane pressure side surface phantom cooling at simulated gas turbine operating conditions (high inlet freestream turbulence level of 16%, exit Mach number of 0.85, and exit Reynolds number of 1.7 × 106). The experimental measurements were conducted at Virginia Tech’s transonic blowdown wind tunnel for two configurations: baseline cascade (HR = 0) and forward-facing step geometry (HR = −0.1). The numerical predictions were performed by solving the steady-state Reynolds-averaged Navier Stokes (RANS) with realizable k–ɛ turbulence model. Based on a double coolant temperature model, a novel numerical method for the predictions of adiabatic wall film cooling effectiveness was proposed. The qualitative and quantitative comparisons between the numerical prediction results and experimental data are provided in this paper. The results indicate that this method can reliably predict endwall film cooling distributions and vane pressure side surface phantom cooling distributions, and significantly reduce (more than 50%) numerical prediction errors. The effects of upstream step geometries were numerically studied at the design flow conditions (BR = 2.5, DR = 1.2), by quantizing the endwall film cooling effectiveness, endwall net heat flux reduction (NHFR), vane pressure side surface phantom cooling effectiveness, and total pressure loss coefficients (TPLC) for various upstream step heights: three forward-facing step heights (HR = −0.16, −0.1, and −0.06), a baseline cascade (HR = 0), and four backward-facing step heights (HR = 0.06, 0.1, 0.136, and 0.2). Results show the upstream forward-facing step geometry is beneficial for enhancing endwall film cooling (maximum enhancement level of 15%), endwall net heat flux reduction (maximum enhancement level of 10%), and vane pressure side surface phantom cooling (maximum enhancement level of 66.7%). Nevertheless, the upstream backward-facing step geometry is pernicious to the endwall film cooling, endwall heat flux reduction, and phantom cooling, and these pernicious effects have sharply deteriorated when the upstream step height is larger than a critical value (HR > 0.136). To be specific, the film cooling effectiveness (from 30% to 80%), net heat flux reduction (up to 78%), and phantom cooling effectiveness (completely disappeared) significantly decrease by introducing an overlarge backward-facing step height HR = 0.2. The upstream step geometries lead to deteriorated aerodynamic performance, though the reduction level may be slight (0.36% at HR = 0.136). These suggest that the overlarge backward-facing misalignment will increase the risk of thermal failure and must be avoided during the component assembly and the operational processes of a gas turbine.