This study considers the multivariable thermodynamic analysis and optimization of transcritical Rankine cycles operating with carbon dioxide as working fluid. Three dependent variables were used as figures of merit: the net power produced by the cycle, and its 1st and 2nd Law efficiencies, all calculated in absolute terms and per unit of global conductance (UA)(Total), where (UA)(Total) accounts for the conductance of all heat exchangers used in the cycle. The key variables were the high pressure of the CO2 within the cycle and the temperature of the heat source, along with four different cycle configurations: (i) a basic power cycle, (ii) a cycle with a recuperator, (iii) a cycle with re-heating and (iv) a cycle with a recuperator and re-heating, namely, combined cycle. The optimization process relied on optimization routines and considered latent and sensible heat sources. This procedure was able to show that while the individually defined figures of merit mostly presented established trends, the normalized figures of merit (i.e., those defined per unit of UA) are highly dependent on the parameters considered and clearly show the existence of optimum values, which are a function of the cycle's configuration, figures of merit considered and operation parameters. (C) 2015 Elsevier Ltd. All rights reserved.