Abstract: To study the effects of intercropping with forage grasses on the composition and diversity of weed communities in an apple orchard in Nyingchi, Tibet, two annual grasses (Avena sativa L. and Vicia sativa L.) and two perennial grasses (Medicago sativa L. and Dactylis glomerata L.) were single sown or mixed sown, with grasses naturally growing after clear tillage used as a control. The weed communities were then surveyed in 2011 and 2022, respectively. A total of 33 weed species belonging to 31 genera and 17 families were found in the experimental plots in the two studied years. The dominant families were Gramineae, Leguminosae, Polygonaceae, and Asteraceae, and the dominant species were Pennisetum flaccidum Griseb., Eragrostis pilosa (L.) Beauv., Galinsoga parviflora Cav., Taraxacum mongolicum Hand.-Mazz., and Persicaria microcephala (D. Don) H. Gross. Intercropping with forage grasses led to a reduction in weed species richness and importance values of the dominant weed species, which were influenced by forage grass species and planting year. Intercropping also resulted in lower species diversity of the weed communities, with both planting year and cropping pattern significantly affecting community diversity and dominance. Weed control efficacy was higher for Gramineae forage compared to Leguminosae forage, and mixed sowing showed better results than single sowing treatment. The above findings suggest that the selection of forage grasses should be based on their growth period, with annual forages being more effective for short-term weed control and perennial forages being more suitable for long-term grass planting in orchards.