The current study investigates the flow of nanofluids in microchannels experimentally and numerically. For this purpose, two microchannels of hydraulic diameters of 211 and 300 μm are used with alumina(45nm)-water nanofluids. The nanofluids with the concentrations 0.25, 0.50 and 1 vol% are used to observe the effect of volume fraction in the present analysis. With regard to the numerical simulation of nanofluids in microchannels, two approaches have been chosen in the current work. First one considers the nanofluids as single phase fluid and applies the mixture rule for evaluating properties for the simulation. The second type of modeling is done using the discrete phase approach which involves Eulerian-Lagrangian considerations. The fluid phase is assumed to be continuous and governed by Navier-Stokes equation. The movement of discrete nanoparticles is determined by the Newton's second law which takes into account the body force, hydrodynamic forces, the Brownian and thermophoresis forces. The predictions are validated against experimental results obtained for nanofluid flow in a chemically etched silicon wafer channel. It is found that the discrete phase modeling is more accurate with regard to the prediction of nanofluids behavior in microchannels, as compared to the single phase model. The results also show the non-uniformity of nanoparticle distribution across the channel cross-section. This non-uniformity in distribution can be attributed to the shear induced particle migration. This can also be the reason for the difference in pressure drop and heat transfer from the single phase model. The pressure drop with 0.25 and 0.5 vol% of alumina is more or less same as that of water and thus, makes it a suitable cooling liquid. However, an enhancement in heat transfer is observed from the computational predictions. © 2010 by ASME.