An LCA of switchgrass production in the Mediterranean region of Spain is carried out, from a cradle-to-farm gate perspective. Experimental plots were previously established in two locations, Moncofar and Orihuela, providing inventory data for 2010-2013. This allowed for the environmental performance to be evaluated throughout a 4-year cycle, considering different sources of variability. The functional unit is 1 t of switchgrass (dry basis) for electricity generation. Besides typical impact categories, blue and green water consumption impacts are also addressed by using watershed characterization factors. In 2010, the production in Orihuela is more input-intensive than in Moncofar, while the biomass yield is lower, causing greater impacts (959.4 vs. 95.9 kg of CO2eq•t-1 for climate change or 19.1 vs. 4.7 kg of Fe-eq•t-1 for metal depletion). In the subsequent years, the yields are higher in Orihuela, hence Moncofar performs worse for some specific impact categories (43.9 vs. 28.5 CO2-eq•t-1 or 8.0 vs. 2.4 kg of Feeq•t-1 in 2011). Due to larger irrigation doses in Orihuela, the blue water impact is always higher than that in Moncofar (on average, 1243 vs. 277 m 3 ecosystem-eq. water•t-1). The green water impact is greater in the latter, except for the first year (on average, 6.9 vs. 53.0 m 3 ecosystem-eq. water•t-1). Overall, both locations deliver sufficient greenhouse gas savings throughout the life cycle, in compliance with the EU requirements for bio-electricity production. However, results show that ad hoc decisions on crop management are critical to the environmental impact, evidencing the importance of considering a multi-year LCA approach.