While known to be superior coolants in stand alone conditions, some scepticism exists with respect to the hydrodynamic and thermodynamic performance of nanofluids in real life applications. The present work employs theoretical investigations (supported by simulation results) on the entropy generation characteristics in parallel microchannel cooling systems (PMCS) employing water and nanofluid as working fluids. Alumina-water nanofluid of different concentrations and PMCS of three different configurations, viz. U, I and Z have been employed for the present study. In order to shed more clarity onto the real thermodynamic performance of nanofluids, an Eulerian-Lagrangian discrete phase approach (DPM) has also been used to model nanofluids alongside the conventional effective property approach (EPM). The thermodynamic performance of twin component nanofluid model in PMCS over the base fluid and single component counterpart has been investigated in view of flow friction generated entropy and heat transfer generated entropy. To quantify thermodynamic irreversibility of nanofluids in PMCS due to heat transfer, the Bejan number has been employed. The entropy generation due to particle migration effects reveal that the effective property model overestimates the entropy generation in case of nanofluids and essentially nanofluids generate lesser degree of entropy than estimated by use of simplistic models. The Bejan number analysis reveals that although hydrodynamically inferior to water, nanofluids are thermodynamically superior fluids when employed as coolants in complex microscale flow systems. The article sheds insight onto the entropy generation behaviour of such dispersed system flows with respect to flow and heat transfer characteristics such as particle concentration, flow Reynolds number, and heat load for proper design of such systems. (C) 2017 Elsevier Ltd. All rights reserved.