Abstract: Surface contamination on aircraft has a significant impact on aerodynamic performance, particularly in terms of laminar flow control, boundary-layer transition, and drag reduction. The complex mechanisms and diverse distribution patterns of surface contamination pose major challenges for modeling and numerical simulation. Firstly, this article proposes a method of abstracting surface contamination as surface roughness, and various roughness distributions and magnitudes are prescribed empirically to represent different contamination types. Using ANSYS Fluent with a transition model, a helicity Reynolds number–based source term is incorporated to enable the prediction of crossflow-induced transition. Furthermore, an empirical correlation derived from experimental data is introduced to improve the transition model, allowing it to predict transition locations under different surface roughness conditions. The modified model is validated through comparative analysis. Finally, for both uniformly distributed and chordwise normally distributed contaminations, quantitative analyses are conducted to investigate the effects of roughness distribution and height on the transition location over the swept wing. Results show that the improved transition model accurately predicts wall transition considering both crossflow and surface roughness effects. The increase in roughness will induce early transition, and the leading edge roughness plays a dominant role in transition induction. When the maximum roughness of the normal distribution contamination is the same as that of the uniform distribution contamination, the calculated transition position is close (less than 5% chord length). The sensitivity of the transition position to roughness decreases with increasing roughness; There is a critical value of 1.14 for uniformly distributed roughness height, below which the transition position varies greatly with roughness, and vice versa, the change tends to be gentle.Surface contamination significantly affects the transition location of swept wings by altering the distribution pattern and characteristic height of roughness. In practical aerodynamic design and maintenance, special attention must be paid to the control of contamination in the leading edge region, with differentiated surface treatment strategies implemented for different roughness-sensitive zones. The above results demonstrate that the improved transition model proposed in this study can effectively predict wall transition locations under crossflow and roughness conditions.