The present paper focuses on analytical and numerical study on using nanofluids as coolant of a microchannel heat sink. The nanofluid studied in this paper is made from copper oxide (CuO) and water. Two common analytical approaches are used: the Fin model and the porous media approach. The Fin model is based on the assumption of uniform fluid temperature in the direction normal to the fluid flow which Causes inaccuracy in the predictions of this approach. On the other hand, modified Darcy equation for the fluid and two-equation model for heat transfer between fluid and solid sections are employed in porous media approach. In addition, to deal with nanofluid heat transfer. a model based on Brownian-motion of nanoparticles is used. The model evaluates the thermal conductivity of nanofluid considering the thermal boundary resistance, nanoparticle diameter, volume fraction and the fluid temperature. This model is included in heat transfer equations of the two approaches. and the velocity profile is obtained analytically considering the transport properties of nanofluids. Firstly, the effects of particle volume fraction and Brownian-Reynolds number on temperature distribution and overall heat transfer coefficient are investigated. After that, the influence of different channel aspect ratios and porosities are studied. Both approaches are considered and compared in details. Furthermore, an optimum aspect ratio is found to minimize the friction factor in different Reynolds numbers. (C) 2009 Elsevier Ltd. All rights reserved.