This work aimed (i) at characterizing the impact of planting density and pedoclimatic conditions on the physiology of poplar and willow dedicated to biomass production, (ii) at evaluating the genetic variability of productivity-related traits, water-use efficiency (WUE) and nitrogen-use efficiency (NUE) in a perspective of plant breeding, and (iii) at assessing the relationships among these traits. Four experimental plantations were established in northern France. Water-use efficiency was indirectly estimated at leaf level through bulk leaf carbon isotope discrimination (Δ13C) and NUE was calculated at plant level as the ratio between the aerial biomass and the stem nitrogen content. Our results clearly show that increasing planting density affects biomass production and WUE depending on site characteristics. At the most favorable site for growth, trees grown at higher density displayed taller stems, lower stem circumference, and higher Δ13C, than trees grown at lower density indicating that increased tree density mainly accentuated competition for light. Under less favorable conditions in terms of water availability and fertility, an increase of planting density involved higher WUE and lower stem circumference, likely because of an increased competition for water. Positive relationships between biomass production and WUE were detected for poplar and willow only under environmental conditions where competition for any resources was reduced. A large canopy is inevitably associated to potentially more important transpiration water losses. These higher water losses at plant level could be compensated at leaf level through a better stomatal regulation. The Genotype x Site interactions indicated that it is difficult to identify good genotypes for a large set of pedoclimatic conditions. The selection of genotypes has to be realized depending on the site conditions. Finally, the absence of antagonism between biomass production, WUE, and NUE suggests the possibility to select genotypes for these three traits independently.