In this study, computational fluid dynamics (CFD) modeling capability of near-wall flow and heat transfer was evaluated against experimental data. Industry-standard wall models for RANS and large-eddy simulation (LES) (law of the wall) were examined against the near-wall flow and heat flux measurements from the transparent combustion chamber (TCC-III) engine. The study shows that the measured, normalized velocity profile does not follow the law of the wall. This wall model, which provides boundary conditions for the simulations, failed to predict the measured velocity profiles away from the wall. LES showed a reasonable prediction in peak heat flux and peak in-cylinder pressure to the experiment, while RANS-heat flux was closer to experimental heat flux but lower in peak pressure. The measurement resolution is higher than that of the simulations, indicating that higher spatial resolution for CFD is needed near the wall to accurately represent the flow and heat transfer. Near-wall mesh refinement was then performed in LES. The wall-normal velocity from the refined mesh case matches better with measurements compared with the wall-parallel velocity. Mesh refinement leads to a normalized velocity profile that matches with measurement in trend only. In addition, the heat flux and its peak value matches well with the experimental heat flux compared with the base mesh.