Background: Forests are increasingly valued for non-timber ecosystem services in place of conventional wood fiber production. Biomass carbon sequestration is one of the key ecosystem services forests are relied upon for mitigating global climate change. However, planning for large-scale tree planting and managing established forest stands for carbon sequestration require careful consideration of the gain in biomass production and tradeoff for other regulatory services. How a tradeoff between forest production and conservation of water resources is shaped by the condition of forest stand and environmental factors remains a question of broad interest in sustainable forest ecosystem management. Methods: We studied the spatiotemporal patterns of net primary productivity (NPP), evapotranspiration (ET), and water use efficiency (WUE), and their relationships with local climatic and forest stand factors over a temperate forest landscape in Changbai Mountain, Northeast China. The time series of spatial data on NPP and ET were extracted from the global remote sensing datasets for the MOD16A3 and MOD17A3 products for the period 2000-2014. The time series of spatial patterns of annual precipitation and annual mean temperature were obtained as grid maps for regional meteorological variables. Stand patches were categorized into the types of conifers, broadleaves, and mixed-wood, as well as age-classes of young, mid-age, near mature, mature, and oldgrowth stands, and by establishment into natural and planted. Information on stands and selective site variables were compiled from the Forest Inventory Datasets of China. Analyses were performed with Arc-GIS. Results: Over the study period of 2000-2014, the landscape-level annual NPP varied between 311.7 and 573.6 gC•m − 2 •a − 1 , ET between 559.9 and 603.0 mm•a − 1 , and WUE between 0.54 and 1.01 gC•m − 2 •mm − 1 . Across the forest landscape, the mean annual NPP varied between 205.0 and 639.4 gC•m − 2 •a − 1 , ET between 441.5 and 784.0 mm•a − 1 , and WUE in the range of 0.46-1.10 gC•m − 2 •mm − 1 . The spatial variations in NPP, ET, and WUE were commonly attributable to forest type, stand age class and density, establishment mode, and temperature variables, with some effects of other selective factors on ET and WUE. The three forest types were significantly (p < 0.05) differentiated in the mean annual NPP, ET, and WUE: the coniferous forests were highest in NPP (505.3 ± 1.4 gC•m − 2 •a − 1 ; n = 1041) and WUE (0.872 ± 0.004 gC•m − 2 •mm − 1 ; n = 1041), and lowest in ET (584.1 ± 1.6 mm•a − 1 ; n = 1041), followed by the mixed-wood forests (NPP: 500.5 ± 0.8 gC•m − 2 •a − 1 ; WUE: 0.856 ± 0.02 gC•m − 2 •mm − 1 ; ET: 589.3 ± 0.9 mm•a − 1 ; n = 2156); whereas the broadleaved forests were lowest in NPP (491.6 ± 0.6 gC•m − 2 •a − 1 ; n = 4428) and WUE (0.832 ± 0.02 gC•m − 2 •mm − 1 ; n = 4428), and highest in ET (594.7 ± 0.6 mm•a − 1 ; n = 4428). The mean annual NPP, ET and WUE increased with stand age typically in coniferous forests, and weakly in mixed-wood forests. The natural sta...