The hydrological processes of mountainous watersheds in inland river basins are complicated.It is absolutely significant to quantify mountainous runoff for social,economic and ecological purposes.This paper takes the mountainous watershed of the Heihe Mainstream River as a study area to simulate the hydrological processes of mountainous watersheds in inland river basins by using the soil and water assessment tool(SWAT)model.SWAT simulation results show that both the Nash–Sutcliffe efficiency and the determination coefficient values of the calibration period(January 1995 to December 2002)and validation period(January 2002 to December 2009)are higher than 0.90,and the percent bias is controlled within±5%,indicating that the simulation results are satisfactory.According to the SWAT performance,we discussed the yearly and monthly variation trends of the mountainous runoff and the runoff components.The results show that from 1996 to 2009,an indistinctive rising trend was observed for the yearly mountainous runoff,which is mainly recharged by lateral flow,and followed by shallow groundwater runoff and surface runoff.The monthly variation demonstrates that the mountainous runoff decreases slightly from May to July,contrary to other months.The mountainous runoff is mainly recharged by shallow groundwater runoff in January,February,and from October to December,by surface runoff in March and April,and by lateral flow from May to September. The hydrological processes of mountainous watersheds in inland river basins are complicated. It is absolutely significant to quantify mountainous runoff for social, economic and ecological purposes. This paper takes the mountainous watershed of the Heihe Mainstream River as a study area to simulate the hydrological processes of mountainous watersheds in inland river basins by using the soil and water assessment tool (SWAT) model. SWAT simulation results show that both both Nash-Sutcliffe efficiency and the determination coefficient values ​​of the calibration period (January 1995 to December 2002) and validation period ( January 2002 to December 2009) are higher than 0.90, and the percent bias is controlled within ± 5%, indicating that the simulation results are satisfactory. According to the SWAT performance, we discussed the yearly and monthly variation trends of the mountainous runoff and the runoff components. The results show that from 1996 to 2009, an indistinctive rising trend was observed for the yearly mountainous runoff, which is mainly recharged by lateral flow, and followed by shallow groundwater runoff and surface runoff. The monthly variation demonstrates that the mountainous runoff decreases slightly from May to July, contrary to other months. groundwater runoff in January, February, and from October to December, by surface runoff in March and April, and by lateral flow from May to September.