Poplar (Populus spp.) and willow (Salix spp.) short rotation coppice (SRC) are attractive feedstock for conversion to renewable electricity. Site managers typically optimize biomass production at their sites. However, maximum biomass production does not necessarily equate an optimal CO2 balance, water use and energy production. This is because many operational actions consume water and energy and emit CO2, either on-site or off-site. Coupling a land surface model (ORCHIDEE-SRC) with life cycle assessment enabled us to determine the optimal management for SRC in Belgium. We simulated 120 different management scenarios for each of two well-studied Belgian SRC sites (i.e. Boom and Lochristi). Simulated soil carbon changes suggested substantial carbon losses of 20-30 Mg ha(-1) over a time period of 20 years, which were within observation-based uncertainty bounds. Results showed that in Belgium, which has a temperate maritime climate, optimal management of SRC has a rotation cycle of two years without irrigation. Energy inputs for this optimal management were 5.2 GJ ha(-1) yr(-1) for the Boom site and 5.3 GJ ha(-1) yr(-1) for the Lochristi site, while the biomass yields at Boom and Lochristi were 9.0 Mg ha(-1) yr(-1) and 9.4 Mg ha(-1) yr(-1), respectively. The energy ratio (i.e., ratio of bioelectricity output to cumulative energy input) for this optimal management was 12, on average. Planting density turned out to be unimportant, while rotation length turned out to be most important to obtain the highest energy ratio and still maintain high biomass yield. Scenarios with high energy-input generated more bioenergy outputs, but the energy gains did not compensate for the increased energy inputs. Reductions in energy consumption per unit of bioenergy output should target the agricultural stage since it accounted for the largest energy share in the production chain.