[Objective] The spatial-temporal variation characteristics of water yield services in Harbin City, Heilongjiang Province from 2000 to 2020 were determined, and the driving mechanism of the spatial-temporal differentiation characteristics of Harbin City's water yield was analyzed in order to provide a scientific basis for the management and sustainable development of water resources in Harbin City. [Methods] The study was conducted in Harbin City, a typical cold land city. The spatial-temporal variation characteristics of water yield in 2000, 2010, and 2020 were analyzed based on the InVEST model's water yield module, and the spatial and temporal evolution of water yield services in Harbin City was determined by using a parameter optimal geographic detector. [Results] ① The water output of Harbin City from 2000 to 2020 increased from 9.68×10⁹ m³ in 2000 to 2.23×10¹⁰ m³ in 2020. The spatial distribution pattern of water yield in different years was basically similar. The overall spatial distribution of water yield was characterized as "higher in the east and lower in the west". ② There was a strong positive spatial correlation in the distribution of water yield in the study area, with the main types being low-low aggregation and high-high aggregation. The proportions of the two areas relative to the city's total area showed a downward trend from 2000 to 2020. ③ The influence of each driving factor on water yield exhibited obvious spatial heterogeneity. Actual evapotranspiration and land use type were the main driving factors of the economic quality development zones. In the nature-dominated ecological barrier area, the comprehensive driving forces of actual evapotranspiration and land use type were far less than in the economic quality development area. [Conclusion] Water yield assessment was closely related to ecosystem services and ecological product value, and serves as the basis of water conservation research. The temporal and spatial changes of water production in Harbin City from 2000 to 2020 were significant, and meteorological and land use factors were the main driving factors.