A numerical investigation is performed into the natural convection heat transfer characteristics and entropy generation of water-based nanofluids in an enclosure bounded by wavy vertical walls and flat upper and lower surfaces. In performing the analysis, it is assumed that the left wall is heated by a constant heat flux while the right wall is maintained at a constant low temperature. In addition, the upper and lower walls are both assumed to be insulated. The analysis considers three different nanofluids, namely Cu-water, Al2O3-water and TiO2-water. The governing equations are modeled using the Boussinesq approximation and are solved using the finite-volume numerical method. The analysis examines the effects of the nanoparticle volume fraction, the type of nanofluid, the Rayleigh number and the wavy surface geometry parameters on the mean Nusselt number and total entropy generation. The results show that for all values of the Rayleigh number considered in the present study (Ra = 10(4)similar to 10(6)), the mean Nusselt number increases and the total entropy generation reduces as the volume fraction of nanoparticles increases. In addition, it is shown that the Cu-water nanofluid yields the best heat transfer performance and the lowest total entropy generation of the three nanofluids. Finally, it is shown that for a given nanofluid, the mean Nusselt number can be maximized and the total entropy generation minimized via an appropriate tuning of the wavy surface geometry parameters. Overall, the results presented in this study provide a useful source of reference for enhancing the natural convection heat transfer performance in wavy-wall enclosures while simultaneously reducing the entropy generation. (C) 2013 Elsevier Ltd. All rights reserved.