Inland river plays an important role in the social and economic development and regional water resources utilization in semi-arid plateau areas of China.To recognize and simulate its evolution characteristics is very significant for the basin's sustainable development.In this paper,Balager river basin was selected as the study object.Based on the hydrological,meteorological, and socio-economic data in 1959-2015,we analyzed the characteristics of the runoff series by the double mass curve method, the cumulative departure method,the improved M-K trend test,and the wavelet transform method.We used statistical correlation test to select the main input factors for runoff evolution simulation,and realized the simulation of annual and seasonal runoff series evolution by BP artificial network.The results showed that the runoff decreased significantly at the annual and seasonal scale.The abrupt change of runoff occurred in 1998,before which the runoff was mainly influenced by climate change,while after which by both human activities and climate change.There was a 6-year period and a 30-year period for the annual and dryseason runoff,yet the wet-season runoff showed no obvious periodic change.The impact factors significantly correlated with runoff included precipitation quantity,wet days,number of precipitation events,evaporation,relative humidity,and all indicators of human activity.The improved BP neural network can well simulate the evolution of runoff in this basin when the distributed hydrological simulation cannot be realized.