<p>The configuration of farmland shelterbelts plays a crucial role in determining their shelter effects. To optimize the configuration patterns of farmland shelterbelts in Xinjiang, wind tunnel simulation experiments were conducted on six mixed tree–shrub shelterbelt configurations at six different angles between the prevailing wind direction and the main shelterbelt orientation. The results revealed that at 90°, Type VI exhibited the strongest windbreak capacity, followed by Type II, whereas Type I performed the worst, with average shelter efficiencies of 45.5%, 43.3%, and 33.6%, respectively. All shelterbelts formed strong deceleration zones within horizontal distances of 1 ~ 3<i>H</i>, with gradual airflow recovery and stabilization occurring at 5 ~ 15<i>H</i>. As the angle between the prevailing wind direction and the main shelterbelt increased, the protective coverage gradually expanded. For Type II and VI shelterbelts at large angles (≥ 60°), significant differences in average wind speed were observed per 15-degree increment, whereas no significant differences were found at small angles (≤ 45°). Therefore, when constructing farmland shelterbelts, priority should be given to the Type II and VI configurations, and an angle deviation of 0–30 degrees between the prevailing wind direction and the main shelterbelt orientation is acceptable. Exceeding this range will reduce the windbreak effectiveness of farmland shelterbelts.</p>