A multi-objective function F(X) by the incorporation of single-objective functions, i.e., net power output W-net, exergy drop of the exhaust gas from inlet to outlet Delta E-g, total exergy destruction rate I and system total cost C-2013, has been put forward as the subcritical organic Rankine cycle (ORC) performance indicator and solved with the method of linear weighted evaluation function. The optimization of the evaporation temperature T-e and condensation temperature T-c of subcritical ORC has been carried out by the use of pure working fluids R600a, R245fa, R601a, R601, R123, and non-azeotropic mixed working fluids R600a/R601a, R245fa/R601a, R245fa/R601, R600a/R245fa. The results reveal that there exist optimal evaporation temperature T-e,T-opt and condensation temperature T-c,T-opt minimizing the multi-objective function F(X); while T-e,T-opt and T-c,T-opt can not exactly exist when choosing single-objective functions, i.e., W-net, I, C-2013 and Delta E-g as the performance indicators. Although T-e,T-opt and T-c,T-opt vary with different working fluids, the ORC performance of mixed working fluids is not always better than that of pure working fluids. The multi-objective optimization of ORC shows superiority to the single-objective optimization, owning to its pursuit of the best comprehensive performance. (C) 2015 Elsevier Ltd. All rights reserved.